Analysis of immune biomarkers after MCS device implantation

load("data_combined.RData")
df$`Device Type` <- relevel(df$`Device Type`, ref = "HMII")
bc <- 14:42
cyt <- 43:80
bcellcyto <- c(bc,cyt)
groups <- make.names(c("AgeGreater60", 
                       "Sex",
                       "LowIntermacs",
                       "RVAD", 
                       "Sensitized",
                       "VAD Indication", 
                       "Survival"
                       ))
FDRcutoff <- 0.1
pcutoff <- 0.05

# HMII wide
df.HMII <- droplevels(subset(df, df$`Device Type` == "HMII"))

suppressMessages(require(knitr, quietly = T))
suppressMessages(require(kableExtra, quietly = T))
suppressMessages(require(rmarkdown, quietly = T))
suppressMessages(require(plyr, quietly = T))
suppressMessages(require(dplyr, quietly = T))
suppressMessages(require(parallel, quietly = TRUE))
suppressMessages(require(ggsci, quietly = TRUE))
suppressMessages(require(ggplot2, quietly = T))
suppressMessages(require(GGally, quietly = T))
suppressMessages(require(reshape2, quietly = T))
suppressMessages(require(stringr, quietly = T))
suppressMessages(require(lmerTest, quietly = TRUE))
suppressMessages(require(car, quietly = TRUE))
suppressMessages(require(fdrtool, quietly = TRUE))
suppressMessages(require(pheatmap, quietly = TRUE))
suppressMessages(require(RColorBrewer, quietly = TRUE))
require(factoextra, quietly = T)
require(NbClust, quietly = T)
require(gridExtra)

# Efron's double standardization
double_standardize <- function(x, niter = 1000) {
    for(i in 1:niter) x <- t(scale(t(scale(x))))
    return(as.data.frame(x))
}

# Error bars
invisible(suppressMessages(require(Hmisc, quietly = T)))
stat_sum_df <- function(fun, geom="errorbar", ...) {
    stat_summary(fun.data = fun, geom = geom, width = 1, ...)
}

# function that converts pvalues to star indicators
p2stars <- function(p){
    if(p >= 0.1) s <- ""
    if(p < 0.1 & p >= 0.05) s <- "."
    if(p < 0.05 & p >= 0.01) s <- "*"
    if(p < 0.01 & p >= 0.001) s <- "**"
    if(p < 0.001 & p >= 0.0001) s <- "***"
    if(p < 0.0001 & p >= 0) s <- "****"
    return(s)
}

HeartMate-II analysis

20 heart-failure patients receiving the Heartmate-II mechanical circulatory support device (MCSD) were sampled at 7 timepoints after implantation. Each sample consisted of 67 biomarker measurements – 29 B-cell markers and 38 cytokine markers. Additionally, each patient was associated with 7 categorical variables, such as age, sex, interMACS score, and survival. Due to practical limitations, not all samples were complete; after accounting for missing data, there are a total of 105 datapoints.

Specific aims

• Our primary aim was to identify if any of the 67 biomarkers are associated with any of the 7 categorical variables, at a false discovery rate of 10%. For this, we used a linear mixed-effect model.

• We also sought to identify any major temporal patterns in the biomarker panel, using clustering.

Methods

Dataset

Data was collected from patients who underwent MCSD implantation at a single university medical center. Peripheral blood draws for cytokine and B-cell expression assays were performed at days 0,1,3,5,8,14,21.

• Cytokine and chemokine concentrations from plasma samples were assayed using the 38-multiplex MILLIPLEX Human Cytokine Chemokine Panel I.

• HLA class I and II single antigen Luminex antibody profiles were performed via flow cytometry on available samples. Allosensitization was defined as HLA antibody production (mean fluorescence intensity > 5000) during the MCSD course.

• B-cell multiparameter immunophenotypes were performed by staining peripheral blood mononuclear cells for surface markers using fluorochrome-tagged antibodies against CD3, CD5, CD11b, CD 19, CD24, CD27, CD38, CD268, IgD, IgM, and IgG (One Lambda, Inc.).

missing.ix <- Reduce(intersect, apply(df.HMII[,bcellcyto], 2, function(x) which(is.na(x))))
df.raw <- df.HMII[-missing.ix,]
df.raw <- df.raw[order(df.raw$PatientID),]
rownames(df.raw) <- 1:nrow(df.raw)
kable(df.raw, 
      digits = 3,
      row.names = T,
      caption = "HeartMate-II Raw Data"
) %>%
    kable_styling(bootstrap_options = c("striped", 
                                        "hover", 
                                        "condensed",
                                        "responsive"),
                  font_size = 12) %>%
    scroll_box(width = "100%", height = "300px") 

HeartMate-II Raw Data

	PatientID	Time	Age	AgeGreater60	Sex	LowIntermacs	InterMACS	RVAD	Sensitized	VAD Indication	Device Type	Outcome	Survival	num Total PBMC	num lymph	lymph	live lymph	CD3 of live lymph	CD19 of live lymph	CD19+CD27-	CD19+CD27+	CD27+38++plasma blasts	CD27-38++ transitional	CD27-IgD+ mature naive	CD27+IgD- switched memory	CD27-IgD- switched memory	CD27+IgD+ unswitched memory	CD27+IgD-IgM+ switched memory	CD27+IgD+IgM+ nonswitched memory	CD19+27+IgG+IgM- memory	CD19+24dim38dim naive mature	CD19+24+38++transitional	CD19CD24hiCD38-memory	CD19+27-38+CD5+transitionals	CD19+CD268+	CD268 of +27-38++transitional	CD19+CD11b+	CD19+CD5+	CD19+CD27+CD24hi	CD19+CD5+CD24hi	CD19+CD5+CD11b+	CD19+27+IgD-38++IgG ASC	IL-12(p40)	IL-12(p70)	IFN-g	TNF-a	TNF-b	IL-4	IL-5	IL-9	IL-10	IL-13	IL-17A	IL-1a	IL-1b	IL-2	IL-3	IL-6	IL-15	TGF-a	IFN-a2	IL-8	GRO	Eotaxin	MDC	IP-10	MCP-1	MCP-3	Fractalkine	MIP-1a	MIP-1b	GM-CSF	IL-7	G-CSF	VEGF	EGF	FGF-2	Flt-3L	IL-1RA	sCD40L
1	1	0	65	older	Male	High	2	No	NA	BTT	HMII	Alive s/p OHT	alive	169154	35496	20.98	99.53	25.37	19.82	86.02	13.98	1.26	2.54	57.81	11.74	28.06	2.39	21.74	14.36	0.72	81.05	2.86	13.60	0.58	95.22	84.83	10.47	4.33	8.50	1.39	3.13	1.09	1.71	2.16	124.000	20.926	2.72	2.500	1.010	7.273	4.617	1.760	21.20	277.000	3.483	9.878	1.160	12.334	1.090	5.71	2.18	45.38	126.00	119.000	525.00	1174	392	3.03	27.25	2.030	47.654	60.843	1.268	33.267	486.000	54.27	6.05	1.92	4.47	793.00
2	1	1	65	older	Male	High	2	No	NA	BTT	HMII	Alive s/p OHT	alive	63082	9915	15.72	99.26	32.45	26.26	90.06	9.94	0.89	2.09	54.27	7.70	35.71	2.32	31.10	20.47	1.57	84.06	2.59	10.79	0.26	97.02	92.59	7.50	3.13	6.42	1.32	2.44	1.49	1.71	1.71	156.000	24.857	2.72	2.500	1.010	1.270	48.726	1.760	29.97	388.000	1.350	1.952	1.160	109.000	5.715	3.89	3.18	102.00	191.00	95.541	463.00	1079	552	3.03	35.53	2.030	62.089	12.293	2.005	109.000	509.000	64.48	26.13	1.92	19.50	841.00
3	1	3	65	older	Male	High	2	No	NA	BTT	HMII	Alive s/p OHT	alive	75921	21721	28.61	99.31	24.86	33.01	86.38	13.62	1.54	1.26	45.74	10.98	40.44	2.84	31.80	18.04	1.15	86.21	4.82	6.56	0.24	90.41	73.33	10.00	7.25	10.12	3.22	5.67	2.05	32.32	14.40	259.000	28.330	14.74	12.770	3.369	5.615	28.055	7.406	54.61	438.000	3.867	8.638	3.690	57.616	12.778	5.85	56.84	121.00	193.00	186.000	510.00	1365	464	35.08	152.00	2.030	87.775	44.032	10.100	81.038	687.000	70.06	117.00	1.92	151.00	801.00
4	1	5	65	older	Male	High	2	No	NA	BTT	HMII	Alive s/p OHT	alive	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	1.71	5.07	134.000	41.230	2.72	2.813	1.809	4.930	23.012	1.760	26.38	316.000	1.219	1.230	2.735	15.579	5.021	2.73	25.80	78.31	231.00	213.000	462.00	2053	583	4.14	104.00	2.030	55.375	49.222	6.820	29.004	414.000	61.03	43.25	1.92	64.51	1230.00
5	1	8	65	older	Male	High	2	No	NA	BTT	HMII	Alive s/p OHT	alive	213808	36002	16.84	99.19	35.95	14.55	67.94	32.06	4.06	2.17	27.92	28.92	39.64	3.52	20.31	9.10	1.33	68.69	7.10	20.42	1.88	87.38	44.25	10.78	11.12	19.15	5.54	7.35	1.07	1.71	3.83	228.000	23.577	2.81	2.500	1.666	6.792	39.103	1.760	43.65	450.000	2.412	6.904	2.197	19.163	2.445	3.75	19.72	56.79	134.00	184.000	496.00	1214	430	3.03	90.07	2.030	72.549	44.681	4.448	51.520	735.000	78.74	94.11	1.92	133.00	912.00
6	1	21	65	older	Male	High	2	No	NA	BTT	HMII	Alive s/p OHT	alive	106970	31867	29.79	99.18	42.36	15.02	84.69	15.31	1.94	2.11	55.41	13.14	29.13	2.32	20.22	12.88	2.08	83.21	5.20	9.50	1.10	94.44	73.00	9.04	6.99	9.22	2.80	4.78	1.44	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
7	3	0	81	older	Male	Low	3	No	NA	DT	HMII	Died	dead	119377	77509	64.93	99.61	58.50	7.99	62.67	37.33	9.10	4.82	36.73	31.90	25.62	5.74	11.47	8.95	0.40	69.54	7.69	9.41	2.61	76.44	57.91	23.29	20.76	15.39	5.42	14.90	2.60	6.39	6.65	11.372	21.497	3.51	2.320	6.303	1.590	6.039	2.111	5.18	4.727	1.920	1.450	2.558	38.074	5.729	2.57	19.65	32.02	123.00	133.000	380.00	1899	1054	26.14	112.00	2.230	32.026	20.483	7.783	58.902	70.513	7.02	95.30	0.39	89.23	624.00
8	3	1	81	older	Male	Low	3	No	NA	DT	HMII	Died	dead	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	41.70	5.35	39.418	40.838	13.66	17.416	2.457	4.050	57.842	9.006	7.14	46.122	4.879	6.666	3.299	119.000	26.234	2.86	67.14	112.00	206.00	180.000	411.00	1271	4441	35.46	169.00	33.419	100.000	54.262	5.253	43.646	118.000	24.49	79.63	0.39	224.00	1714.00
9	3	3	81	older	Male	Low	3	No	NA	DT	HMII	Died	dead	96940	57658	59.48	99.70	55.52	9.21	72.15	27.85	3.85	3.82	48.41	21.75	23.49	6.35	12.10	16.04	0.28	75.30	3.72	6.65	2.11	80.52	67.33	15.80	15.23	6.71	1.32	8.86	0.26	17.15	6.00	20.672	15.342	7.00	5.509	7.681	2.040	8.932	3.799	4.45	17.129	2.699	3.725	3.108	25.794	15.326	2.57	74.20	34.62	85.04	162.000	348.00	897	1361	21.65	102.00	6.821	27.057	24.672	5.253	34.257	70.513	13.56	63.96	0.39	118.00	1014.00
10	3	5	81	older	Male	Low	3	No	NA	DT	HMII	Died	dead	154373	73969	47.92	99.65	45.18	8.77	67.36	32.64	6.59	2.77	39.10	26.28	27.95	6.67	17.89	15.71	0.39	71.57	6.26	9.88	1.78	81.86	72.07	20.12	20.17	14.80	5.99	14.65	3.71	4.28	2.29	12.657	16.490	2.31	2.320	6.816	1.590	6.845	1.520	4.74	2.120	1.920	1.450	1.980	99.474	11.334	1.58	47.29	61.82	48.64	192.000	320.00	738	1071	10.05	46.41	2.230	21.777	11.020	1.520	37.403	27.750	2.89	23.91	0.39	52.69	361.00
11	3	8	81	older	Male	Low	3	No	NA	DT	HMII	Died	dead	155610	63375	40.73	99.51	38.42	12.40	71.69	28.31	4.36	2.62	37.82	23.43	33.50	5.24	17.37	14.92	0.60	65.04	2.80	11.23	2.54	83.55	37.56	15.55	15.45	9.55	2.66	10.91	0.28	8.54	4.12	17.955	15.342	3.51	2.320	6.816	1.590	6.439	1.605	5.62	13.939	2.314	1.782	2.042	79.869	8.070	2.42	84.63	34.26	38.88	162.000	335.00	822	1054	16.46	84.38	2.980	19.937	20.483	2.734	21.566	90.169	10.50	41.25	0.39	77.29	562.00
12	3	14	81	older	Male	Low	3	No	NA	DT	HMII	Died	dead	244245	115109	47.13	99.47	47.40	9.37	67.50	32.50	3.99	2.08	50.08	26.01	17.22	6.68	15.34	15.25	0.23	77.86	3.21	10.49	1.48	82.13	59.19	21.45	19.77	14.45	2.57	14.94	0.94	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
13	4	1	58	younger	Male	Low	3	Yes	No	BTT	HMII	Alive s/p OHT	alive	246811	86529	35.06	93.49	67.96	17.48	81.66	18.34	2.05	6.85	51.33	13.69	29.91	5.08	15.68	5.94	0.19	61.95	0.76	13.51	4.07	92.62	34.06	6.90	12.07	8.00	2.54	3.85	0.26	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
14	4	3	58	younger	Male	Low	3	Yes	No	BTT	HMII	Alive s/p OHT	alive	406771	90343	22.21	97.34	49.28	17.02	75.04	24.96	4.47	18.36	28.45	21.86	46.07	3.62	24.97	4.33	0.13	47.41	1.64	19.70	18.15	72.76	8.08	25.69	28.35	13.95	4.66	19.20	0.41	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
15	4	21	58	younger	Male	Low	3	Yes	No	BTT	HMII	Alive s/p OHT	alive	330191	81636	24.72	94.57	45.14	14.70	70.51	29.49	8.94	16.23	38.92	23.53	30.83	6.72	23.93	7.07	0.18	46.63	2.35	16.12	8.59	74.45	39.90	24.19	21.18	12.62	3.26	12.93	0.38	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
16	7	0	65	older	Female	High	2	No	Yes	BTT	HMII	Alive s/p OHT	alive	1073919	336538	31.34	100.00	63.29	14.80	93.05	6.95	0.73	8.05	75.49	4.55	19.16	0.80	7.25	8.78	28.04	88.72	2.51	5.18	1.02	90.49	68.18	2.32	4.09	1.69	3.48	1.27	5.78	103.00	34.40	73.796	17.487	330.00	74.260	18.401	16.683	51.466	82.515	12.80	150.000	15.178	15.208	9.970	26.235	21.107	8.44	133.00	39.26	415.00	111.000	298.00	443	296	138.00	125.00	10.503	37.240	39.468	13.629	111.000	409.000	77.43	78.89	10.95	301.00	5998.00
17	7	1	65	older	Female	High	2	No	Yes	BTT	HMII	Alive s/p OHT	alive	270301	97229	35.97	91.63	18.43	61.67	97.68	2.32	0.20	5.85	69.45	1.63	28.19	0.73	27.57	21.43	17.05	84.32	7.26	2.73	0.84	99.44	97.60	0.54	2.64	1.51	3.63	0.29	3.40	16.38	6.78	46.507	34.398	28.79	11.752	0.814	4.999	1481.000	8.712	8.92	17.769	2.325	4.279	2.360	63.109	16.386	5.90	34.66	152.00	295.00	55.059	278.00	1037	409	21.48	40.32	6.177	30.898	20.452	2.676	177.000	224.000	36.40	47.22	0.92	160.00	5604.00
18	7	3	65	older	Female	High	2	No	Yes	BTT	HMII	Alive s/p OHT	alive	499016	78019	15.63	89.41	25.83	51.38	92.98	7.02	0.32	4.66	51.34	6.31	41.59	0.76	36.49	5.96	22.81	73.48	6.73	7.93	0.33	98.78	96.52	1.31	2.06	5.19	2.68	0.59	2.45	16.38	6.52	31.962	18.537	34.49	16.410	2.113	4.287	78.861	9.272	9.04	17.769	2.902	2.861	3.595	11.983	22.669	9.71	14.43	38.26	183.00	56.470	219.00	307	219	29.76	47.34	5.053	3.891	7.570	2.878	33.464	191.000	16.92	81.03	0.92	99.54	4583.00
19	7	8	65	older	Female	High	2	No	Yes	BTT	HMII	Alive s/p OHT	alive	115825	32513	28.07	91.04	44.61	15.76	80.52	19.48	4.16	7.82	49.91	18.37	30.52	1.20	14.87	3.33	30.19	61.08	8.70	12.56	1.58	89.80	76.71	2.87	6.07	11.12	8.08	1.59	1.17	37.61	21.91	70.846	24.832	57.61	38.448	5.988	8.685	44.610	13.488	17.95	55.857	7.772	7.570	7.238	14.443	23.376	10.40	81.88	77.46	332.00	132.000	332.00	1453	632	38.47	104.00	12.646	40.206	31.261	7.509	79.659	354.000	26.71	180.00	0.92	274.00	7156.00
20	7	21	65	older	Female	High	2	No	Yes	BTT	HMII	Alive s/p OHT	alive	407080	97651	23.99	91.29	66.96	12.20	84.66	15.34	2.88	5.60	61.89	13.73	22.65	1.74	20.16	7.23	29.39	68.49	5.28	10.60	1.17	92.50	54.35	3.48	6.53	9.61	8.34	1.85	2.83	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
21	8	0	43	younger	Male	High	2	Yes	No	BTT	HMII	Alive s/p OHT	alive	515760	266115	51.60	96.30	69.66	8.57	72.22	27.78	6.26	4.77	43.09	26.01	28.26	2.63	16.67	2.67	28.25	50.35	0.41	12.78	4.47	84.76	19.47	9.56	4.55	8.29	0.87	2.18	0.75	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
22	8	8	43	younger	Male	High	2	Yes	No	BTT	HMII	Alive s/p OHT	alive	539521	223872	41.49	96.72	60.02	12.63	80.13	19.87	1.66	2.48	43.56	19.52	35.24	1.68	20.46	3.67	20.54	55.76	0.18	10.80	2.43	87.82	13.25	6.55	5.96	4.65	0.95	2.51	0.65	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
23	8	21	43	younger	Male	High	2	Yes	No	BTT	HMII	Alive s/p OHT	alive	564317	216312	38.33	97.40	55.80	4.83	74.19	25.81	5.35	5.77	37.90	25.72	34.76	1.62	17.99	2.18	31.75	61.61	0.77	13.44	6.97	76.55	6.97	10.34	11.14	4.46	0.94	5.14	2.43	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
24	10	0	43	younger	Male	Low	3	No	No	BTT	HMII	Alive s/p OHT	alive	92221	27988	30.35	85.45	55.78	1.51	85.56	14.44	0.28	0.28	55.28	4.72	39.72	0.28	0.00	1.89	1.89	89.44	0.00	2.78	0.00	15.00	0.00	3.33	0.83	0.28	0.28	3.33	0.00	7.64	8.05	12.160	18.210	2.65	2.810	2.640	2.960	12.830	1.500	2.82	2.930	1.340	0.600	2.050	88.780	4.370	4.15	28.83	39.26	873.00	84.790	386.00	413	361	15.64	102.00	3.080	25.630	12.440	3.660	40.120	112.000	84.03	48.41	2.75	67.22	13502.00
25	10	1	43	younger	Male	Low	3	No	No	BTT	HMII	Alive s/p OHT	alive	273746	46755	17.08	93.25	50.14	7.57	88.27	11.73	0.09	2.12	54.52	3.36	42.00	0.12	8.40	6.87	0.00	92.00	0.12	3.45	0.00	3.94	11.43	1.12	0.64	2.76	0.33	2.15	0.00	1.74	3.20	8.390	13.870	2.65	2.810	2.640	2.960	5.480	1.500	2.63	1.040	0.420	0.020	0.960	56.230	6.110	2.78	12.80	22.38	257.00	109.000	212.00	207	352	2.90	59.32	3.080	10.480	4.350	2.300	35.160	31.280	16.68	40.19	1.22	37.18	3646.00
26	10	3	43	younger	Male	Low	3	No	No	BTT	HMII	Alive s/p OHT	alive	156053	71752	45.98	78.38	35.43	27.12	78.57	21.43	0.24	0.60	58.56	5.80	18.33	17.31	12.67	72.96	0.53	94.06	0.49	5.11	0.28	22.79	27.47	0.81	5.53	0.01	0.29	0.19	0.00	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
27	10	5	43	younger	Male	Low	3	No	No	BTT	HMII	Alive s/p OHT	alive	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	28.26	29.24	25.930	19.130	12.01	9.860	4.910	2.960	12.270	2.300	8.31	8.480	4.420	6.460	5.600	31.690	12.300	6.41	65.76	22.02	605.00	111.000	650.00	279	320	27.98	245.00	8.330	34.600	26.450	10.410	112.000	184.000	48.97	197.00	24.56	128.00	8387.00
28	10	8	43	younger	Male	Low	3	No	No	BTT	HMII	Alive s/p OHT	alive	156053	71752	45.98	78.38	35.43	27.12	78.57	21.43	0.24	0.60	58.56	5.80	18.33	17.31	12.67	72.96	0.53	94.06	0.49	5.11	0.28	22.79	27.47	0.81	5.53	0.01	0.29	0.19	0.00	70.58	65.90	45.390	26.340	33.60	24.140	10.790	5.940	18.220	4.580	16.86	22.040	10.800	17.200	14.210	33.990	19.940	11.61	146.00	23.83	661.00	105.000	765.00	325	270	45.41	414.00	17.810	51.840	49.970	22.030	205.000	266.000	76.21	266.00	48.57	276.00	9997.00
29	13	0	61	older	Male	High	2	No	No	BTT	HMII	Alive s/p OHT	alive	411616	119210	28.96	90.82	73.63	5.88	74.30	25.70	2.80	6.46	57.94	19.20	18.72	4.13	19.10	15.42	15.30	53.51	13.78	25.19	0.21	47.23	53.77	5.75	5.92	22.63	3.46	2.55	0.74	74.33	630.00	1198.000	29.033	285.00	6.710	41.890	282.000	58.746	108.000	242.00	1363.000	17.372	74.166	4.089	381.000	11.334	76.55	183.00	145.00	703.00	339.000	938.00	642	424	191.00	439.00	51.380	409.000	169.000	31.768	254.000	3217.000	299.00	1084.00	546.00	877.00	6887.00
30	13	1	61	older	Male	High	2	No	No	BTT	HMII	Alive s/p OHT	alive	430509	79492	18.46	89.31	52.10	20.45	87.13	12.87	1.23	4.76	58.11	9.15	30.55	2.20	29.86	18.05	10.37	63.63	10.23	17.43	0.06	74.05	72.50	2.36	2.98	10.68	1.72	1.03	2.89	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
31	13	3	61	older	Male	High	2	No	No	BTT	HMII	Alive s/p OHT	alive	395713	119670	30.24	97.90	23.79	6.27	82.17	17.83	0.79	1.99	49.95	13.01	33.91	3.13	28.36	16.83	5.00	56.33	7.13	26.73	0.02	79.50	82.88	1.96	5.55	16.30	3.61	1.88	0.42	4.28	159.00	409.000	41.124	12.27	2.320	4.307	20.356	15.581	16.793	121.00	61.854	2.899	4.794	1.980	49.157	20.020	28.46	32.56	102.00	524.00	415.000	305.00	282	465	41.67	203.00	12.431	121.000	35.967	2.120	73.622	945.000	39.60	451.00	55.87	159.00	676.00
32	13	8	61	older	Male	High	2	No	No	BTT	HMII	Alive s/p OHT	alive	329840	73468	22.27	98.63	70.61	3.73	51.50	48.50	8.32	3.77	33.46	41.29	20.92	4.33	25.19	5.89	24.07	46.40	3.73	34.64	0.07	64.14	28.43	10.35	9.61	38.78	4.62	5.55	2.91	100.00	430.00	1063.000	57.719	245.00	135.000	21.182	84.771	41.826	160.000	229.00	580.000	22.030	72.412	10.670	229.000	25.806	54.08	223.00	143.00	966.00	310.000	777.00	1415	796	217.00	325.00	45.047	411.000	137.000	44.518	266.000	2406.000	193.00	908.00	321.00	732.00	7464.00
33	13	14	61	older	Male	High	2	No	No	BTT	HMII	Alive s/p OHT	alive	350294	88823	25.36	97.63	74.37	4.37	73.35	26.65	4.91	4.83	56.54	21.32	19.84	2.30	34.68	11.25	18.37	61.98	5.78	20.55	0.04	63.91	47.54	6.81	6.17	21.77	2.90	2.74	3.00	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
34	16	0	39	younger	Female	Low	3	Yes	No	BTT	HMII	Died post OHT	dead	52016	30920	59.44	95.77	55.57	24.62	58.73	41.27	2.14	0.97	53.01	5.05	5.45	36.50	9.39	81.29	2.88	81.48	5.65	9.02	0.48	97.75	71.83	0.67	22.89	0.34	1.78	0.49	NA	11.66	5.99	5.180	18.210	2.65	2.810	2.640	2.960	2.600	1.500	3.19	3.790	0.750	1.140	1.270	12.840	1.920	2.78	22.29	13.36	205.00	120.000	433.00	852	235	13.38	108.00	17.980	22.860	7.890	3.660	26.920	114.000	27.36	63.49	6.10	47.28	2960.00
35	16	1	39	younger	Female	Low	3	Yes	No	BTT	HMII	Died post OHT	dead	75516	48801	64.62	92.13	52.90	18.64	70.39	29.61	2.92	2.21	62.18	5.13	7.98	24.70	13.79	69.54	3.19	84.14	3.64	6.36	0.41	94.93	43.78	0.57	15.61	0.23	0.91	0.36	NA	6.69	4.54	3.850	13.310	2.65	2.810	2.640	2.960	15.360	1.500	2.89	2.390	0.640	1.550	1.060	51.270	3.890	2.78	23.91	29.12	139.00	53.210	184.00	503	192	13.38	162.00	30.710	25.630	7.270	3.380	58.260	135.000	10.50	52.34	4.60	111.00	1306.00
36	16	3	39	younger	Female	Low	3	Yes	No	BTT	HMII	Died post OHT	dead	250612	130454	52.05	95.75	60.05	3.29	46.04	53.96	1.34	3.89	33.08	32.47	8.29	26.15	3.56	34.61	0.09	76.71	0.39	13.20	0.28	39.79	14.38	2.26	1.56	10.45	2.89	3.04	0.08	22.58	5.49	6.580	15.690	2.77	2.810	2.640	2.960	19.820	1.500	3.96	4.710	1.750	5.060	1.270	242.000	7.410	6.23	43.70	45.47	205.00	139.000	201.00	614	498	27.61	259.00	37.910	41.540	13.790	4.520	90.330	187.000	20.67	91.18	20.95	94.90	1637.00
37	16	5	39	younger	Female	Low	3	Yes	No	BTT	HMII	Died post OHT	dead	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	36.87	8.59	9.500	14.150	5.49	4.340	2.640	2.960	3.860	1.500	4.92	8.120	2.050	5.620	1.480	49.360	5.860	3.77	51.10	19.90	214.00	142.000	261.00	1163	556	27.24	145.00	37.770	41.540	20.750	5.700	54.980	210.000	28.80	99.93	38.44	100.00	3060.00
38	16	8	39	younger	Female	Low	3	Yes	No	BTT	HMII	Died post OHT	dead	212478	120477	56.70	85.12	48.36	12.51	79.57	20.43	1.35	2.14	58.73	12.87	18.44	9.96	5.63	34.32	0.11	86.74	0.41	6.88	0.15	35.30	53.09	1.23	0.75	2.36	1.04	1.47	NA	39.75	7.52	10.260	14.990	6.70	4.900	2.640	3.060	5.960	1.500	5.57	11.870	2.670	6.610	1.480	49.000	5.860	3.86	55.20	17.87	278.00	105.000	428.00	1180	309	30.12	162.00	41.370	49.350	22.170	5.990	67.210	220.000	34.39	108.00	31.81	119.00	3873.00
39	17	0	70	older	Female	High	2	No	NA	DT	HMII	Alive	alive	86135	70306	81.62	93.71	74.50	1.59	61.74	38.26	1.91	2.67	33.97	14.50	26.81	24.71	20.67	62.47	9.03	73.00	2.00	24.71	1.27	20.23	71.43	0.48	12.79	0.38	3.24	0.48	4.60	2.81	7.26	7.290	173.000	2.65	2.810	2.640	2.960	12.550	1.500	4.28	8.850	0.640	0.770	1.160	2.930	3.890	3.28	27.19	22.83	145.00	70.170	765.00	1478	482	9.57	94.83	94.860	103.000	8.520	3.660	18.060	120.000	19.71	40.19	28.19	51.59	1276.00
40	17	1	70	older	Female	High	2	No	NA	DT	HMII	Alive	alive	81004	41334	51.03	94.06	53.64	2.99	61.70	38.30	3.10	2.75	39.07	13.34	22.12	25.47	12.58	63.11	15.99	78.92	2.93	17.38	3.35	23.24	71.88	0.52	11.70	0.09	1.29	0.52	9.83	7.64	4.08	4.170	41.980	2.65	2.810	2.640	2.960	133.000	1.500	2.63	7.760	0.420	1.770	0.910	98.210	8.740	4.54	15.89	77.14	211.00	64.900	374.00	1714	959	5.00	64.16	21.350	54.290	8.520	2.560	284.000	52.100	11.10	21.92	7.15	51.59	1301.00
41	17	3	70	older	Female	High	2	No	NA	DT	HMII	Alive	alive	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	0.22	3.20	3.220	21.580	2.65	2.810	2.640	2.960	15.210	1.500	2.63	2.520	0.420	0.020	0.680	58.680	7.800	3.77	8.37	31.94	83.16	74.380	388.00	644	432	2.90	49.55	8.660	30.220	4.630	1.790	14.940	45.600	11.69	14.26	1.53	29.96	690.00
42	17	5	70	older	Female	High	2	No	NA	DT	HMII	Alive	alive	90024	39434	43.80	95.32	68.83	5.62	71.59	28.41	0.85	1.66	35.23	13.02	35.80	15.96	22.94	52.37	5.85	76.42	1.14	21.83	0.62	14.54	48.57	0.38	4.69	0.00	0.80	0.38	0.62	0.58	3.63	3.220	25.010	2.65	2.810	2.640	2.960	23.480	1.500	2.63	2.390	0.420	0.100	0.860	31.820	7.670	11.61	15.89	41.91	454.00	91.330	506.00	670	421	3.21	54.44	14.370	37.430	6.660	2.300	33.500	58.260	15.08	19.42	6.10	41.52	1837.00
43	17	8	70	older	Female	High	2	No	NA	DT	HMII	Alive	alive	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	37.61	21.91	44.071	23.633	39.20	9.532	6.428	5.102	32.860	7.062	12.93	20.414	2.167	3.138	5.978	41.851	8.190	4.95	63.21	39.70	328.00	175.000	323.00	808	540	34.88	91.67	13.604	48.125	30.231	5.864	59.949	466.000	81.74	150.00	4.51	195.00	9741.00
44	17	14	70	older	Female	High	2	No	NA	DT	HMII	Alive	alive	106791	80532	75.41	88.82	79.92	1.57	62.32	37.68	1.70	4.02	28.04	22.86	34.11	15.00	23.92	39.41	15.03	61.61	3.66	34.55	1.97	20.36	73.33	0.80	8.57	0.27	1.79	0.98	0.71	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
45	18	0	63	older	Male	High	1	No	NA	BTT	HMII	Died	dead	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	60.38	32.81	186.000	38.833	32.98	21.709	3.051	5.621	37.885	11.091	29.33	189.000	6.338	16.504	4.089	278.000	17.668	7.17	59.99	54.99	695.00	232.000	211.00	828	1145	50.86	107.00	10.251	111.000	31.722	7.153	120.000	533.000	28.75	322.00	2.34	224.00	3524.00
46	18	1	63	older	Male	High	1	No	NA	BTT	HMII	Died	dead	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	63.95	40.14	249.000	28.453	32.98	21.709	3.357	4.711	28.562	11.091	38.39	220.000	6.089	13.032	4.089	124.000	15.114	5.01	59.99	41.74	638.00	289.000	303.00	509	713	59.63	121.00	9.867	123.000	56.352	7.153	82.196	685.000	46.72	384.00	14.39	229.00	5489.00
47	18	3	63	older	Male	High	1	No	NA	BTT	HMII	Died	dead	366713	77282	21.07	98.61	66.55	12.44	93.65	6.35	0.53	0.60	62.22	4.68	32.38	0.72	19.94	9.89	28.23	83.18	0.51	14.38	0.30	94.72	54.39	2.07	11.33	2.86	3.40	1.48	3.58	65.71	24.75	198.000	26.329	40.68	18.837	4.631	4.711	20.178	9.696	29.66	185.000	5.841	11.411	4.293	71.642	12.166	3.92	74.20	35.60	1110.00	222.000	300.00	489	627	50.86	126.00	9.479	104.000	51.467	10.275	77.933	727.000	62.42	422.00	11.03	219.00	9557.42
48	18	5	63	older	Male	High	1	No	NA	BTT	HMII	Died	dead	262251	55344	21.10	99.55	62.37	14.39	93.00	7.00	1.03	0.69	65.71	4.74	28.37	1.17	25.93	17.28	25.93	79.87	1.20	17.05	0.48	97.76	58.18	2.27	11.13	4.12	4.91	1.87	6.10	89.92	70.33	309.000	36.344	103.00	29.010	8.732	7.037	28.329	24.712	57.19	384.000	8.403	25.681	5.347	93.592	18.950	6.40	67.14	51.06	434.00	238.000	328.00	896	536	91.53	134.00	14.161	155.000	92.272	10.890	115.000	873.000	64.87	398.00	65.30	368.00	2247.00
49	18	8	63	older	Male	High	1	No	NA	BTT	HMII	Died	dead	309193	61091	19.76	98.15	56.62	7.24	83.61	16.39	3.27	1.52	58.03	11.90	28.13	1.93	24.65	11.36	1.52	79.53	1.45	13.33	0.97	92.63	25.76	3.50	9.23	7.57	3.73	3.48	0.00	60.38	17.64	183.000	34.233	42.23	20.268	6.645	4.488	29.029	11.795	27.05	177.000	6.089	10.605	3.887	154.000	12.585	3.77	56.39	56.34	1534.00	165.000	446.00	487	478	53.19	121.00	10.759	85.242	37.382	13.310	79.359	700.000	104.00	490.00	12.72	151.00	9557.42
50	19	0	68	older	Male	Low	3	No	NA	DT	HMII	Died	dead	147997	54649	36.93	91.75	79.63	7.62	74.86	25.14	0.60	3.45	62.45	22.87	12.30	2.38	45.98	9.28	13.61	74.33	2.67	21.19	0.84	91.31	90.15	3.58	2.25	21.79	1.36	0.58	0.34	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
51	19	1	68	older	Male	Low	3	No	NA	DT	HMII	Died	dead	177687	56400	31.74	97.13	69.76	12.60	81.83	18.17	0.65	1.61	55.77	16.13	26.02	2.09	53.19	10.70	11.57	82.92	1.38	13.98	0.28	89.83	85.59	3.78	1.85	14.65	1.30	0.68	0.18	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
52	19	3	68	older	Male	Low	3	No	NA	DT	HMII	Died	dead	233155	67140	28.80	98.44	74.89	5.32	81.96	18.04	1.00	1.88	49.03	16.59	32.90	1.48	49.92	7.20	12.83	83.44	1.39	12.98	0.14	89.70	69.70	4.58	1.62	13.89	1.08	1.08	0.17	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
53	19	5	68	older	Male	Low	3	No	NA	DT	HMII	Died	dead	99257	29455	29.68	97.33	62.48	4.73	80.60	19.40	1.84	1.25	45.06	18.36	35.25	1.33	44.61	6.32	12.27	82.89	0.88	13.20	0.09	83.55	35.29	7.23	3.54	14.75	2.29	3.02	1.20	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
54	19	8	68	older	Male	Low	3	No	NA	DT	HMII	Died	dead	76358	18702	24.49	97.32	29.48	11.10	86.29	13.71	2.43	0.50	37.92	12.13	48.32	1.63	46.45	11.70	11.35	83.96	0.20	10.50	0.06	57.82	50.00	9.95	29.60	10.20	0.64	9.36	0.80	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
55	21	0	46	younger	Male	High	2	No	NA	DT	HMII	Alive	alive	226762	25868	11.41	89.41	25.31	12.05	55.06	44.94	11.31	2.69	14.79	42.57	39.95	2.69	19.87	3.88	32.88	68.95	10.48	13.57	0.59	80.62	22.67	31.91	19.38	15.58	10.37	16.55	8.18	2.18	2.45	60.794	34.617	2.31	2.320	1.245	1.590	68.832	1.520	7.45	2.120	1.920	1.450	1.980	130.000	3.544	14.61	5.57	47.27	442.00	52.814	249.00	328	245	3.98	55.48	2.230	19.937	2.737	3.987	21.566	118.000	2.89	23.91	0.39	52.69	1293.00
56	21	3	46	younger	Male	High	2	No	NA	DT	HMII	Alive	alive	244049	39386	16.14	94.94	40.46	13.23	40.93	59.07	13.36	3.17	8.33	54.90	32.28	4.49	18.52	4.21	23.83	55.18	13.87	19.39	1.44	70.53	35.67	33.62	33.47	22.86	15.26	25.81	5.11	2.18	2.71	17.955	26.715	9.57	2.320	1.893	1.590	3.759	1.520	3.89	2.120	1.920	1.450	1.980	59.228	1.623	7.17	8.93	60.31	820.00	73.754	165.00	347	262	3.98	63.62	2.230	18.052	2.737	3.358	15.234	90.169	11.56	77.19	0.39	33.64	2726.00
57	21	5	46	younger	Male	High	2	No	NA	DT	HMII	Alive	alive	344736	250121	72.55	89.83	7.54	3.81	45.84	54.16	3.96	1.53	13.39	52.70	32.18	1.72	67.14	2.92	2.44	74.68	6.81	13.91	0.21	63.33	58.02	33.05	30.94	19.83	13.96	26.74	0.22	2.18	4.12	15.278	24.202	12.27	2.320	2.457	1.590	4.492	1.520	3.07	2.120	1.920	1.450	2.558	43.554	1.623	2.86	23.32	48.13	1082.00	71.954	125.00	260	251	5.90	77.95	2.230	18.052	7.234	7.153	27.925	153.000	29.97	118.00	0.39	52.69	4714.00
58	21	8	46	younger	Male	High	2	No	NA	DT	HMII	Alive	alive	216934	140093	64.58	96.37	4.78	7.22	22.49	77.51	13.49	0.64	7.22	67.61	15.17	10.00	35.88	7.82	17.22	51.28	39.40	6.56	1.33	62.59	58.06	66.67	64.59	53.00	50.26	63.78	4.97	2.18	2.45	24.797	23.042	8.26	2.320	1.625	1.590	3.060	1.520	4.03	2.120	1.920	1.450	1.980	44.329	0.827	2.13	5.57	54.35	370.00	84.966	89.52	348	255	3.98	46.41	2.230	9.990	2.520	2.120	9.035	55.097	2.89	23.91	0.39	27.20	1113.00
59	21	14	46	younger	Male	High	2	No	NA	DT	HMII	Alive	alive	215820	66785	30.94	92.54	17.20	6.01	25.88	74.12	9.15	1.45	3.09	71.72	22.44	2.74	65.43	3.42	0.04	69.38	15.93	8.85	0.74	50.58	24.07	52.78	48.88	32.28	28.19	44.04	0.04	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
60	21	21	46	younger	Male	High	2	No	NA	DT	HMII	Alive	alive	670692	523014	77.98	96.05	5.07	2.64	23.61	76.39	4.11	0.49	4.79	73.53	18.68	3.00	42.40	2.72	4.49	71.95	20.81	5.51	0.45	48.22	44.62	60.13	60.26	39.07	36.82	56.38	1.49	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
61	22	0	75	older	Male	High	2	No	NA	DT	HMII	Alive	alive	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	59.97	33.76	44.886	35.293	48.50	30.282	52.148	11.702	46.900	18.224	10.78	72.616	5.569	5.458	11.016	23.895	17.965	7.15	143.00	43.59	989.00	135.000	584.00	662	925	40.52	163.00	13.354	43.750	59.591	13.629	94.278	449.000	104.00	196.00	0.92	248.00	10443.75
62	22	1	75	older	Male	High	2	No	NA	DT	HMII	Alive	alive	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	6.42	13.61	24.600	30.292	4.17	3.956	10.044	3.491	27.540	3.964	5.69	3.190	0.940	1.805	2.360	69.905	14.225	3.22	64.43	56.18	263.00	47.791	495.00	624	1321	16.57	80.75	6.352	20.019	15.869	3.706	57.971	119.000	6.20	184.00	0.92	49.17	2297.00
63	22	3	75	older	Male	High	2	No	NA	DT	HMII	Alive	alive	121794	28129	23.10	97.05	52.79	16.42	70.08	29.92	0.29	0.54	50.71	21.46	22.96	4.86	34.18	12.51	23.49	39.78	0.31	55.18	0.35	96.01	16.67	1.74	5.27	23.38	2.95	1.00	0.10	2.77	13.90	16.630	13.668	9.16	2.970	3.372	2.078	10.277	6.347	2.66	4.260	1.165	1.090	2.360	36.910	7.340	2.59	60.77	27.10	698.00	100.000	514.00	323	606	25.83	54.27	4.276	23.867	33.318	4.024	31.548	264.000	61.66	124.00	0.92	69.60	10443.75
64	22	5	75	older	Male	High	2	No	NA	DT	HMII	Alive	alive	82157	17295	21.05	99.20	72.54	7.41	68.21	31.79	1.10	0.55	48.47	22.90	23.60	5.04	31.90	13.57	18.10	37.84	0.47	56.73	0.35	92.37	14.29	2.05	7.79	26.20	4.72	1.42	1.02	17.15	11.58	21.356	25.652	9.57	5.509	5.963	3.002	17.410	6.315	3.89	23.573	2.504	1.494	5.132	87.134	12.166	2.86	56.39	33.82	845.00	150.000	484.00	556	814	19.17	130.00	6.685	27.906	20.483	10.890	40.533	185.000	70.35	99.39	2.34	118.00	7899.00
65	22	8	75	older	Male	High	2	No	NA	DT	HMII	Alive	alive	154810	27067	17.48	98.04	72.18	5.03	74.91	25.09	0.90	1.65	58.05	14.76	20.82	6.37	23.43	25.71	22.57	44.94	0.60	49.36	0.71	86.59	22.73	3.30	7.64	20.15	4.87	1.72	0.00	10.19	18.30	30.677	25.655	16.47	9.100	6.693	4.896	29.159	3.022	8.07	3.688	1.295	1.557	8.079	68.056	13.360	5.90	58.29	31.79	671.00	47.791	696.00	506	704	18.84	102.00	12.748	17.993	11.877	5.864	44.162	198.000	49.71	115.00	1.39	114.00	10443.75
66	22	21	75	older	Male	High	2	No	NA	DT	HMII	Alive	alive	197936	15446	7.80	95.86	50.35	5.44	79.13	20.87	3.98	3.11	52.30	15.90	29.81	1.99	20.23	5.78	24.86	40.62	1.12	50.06	1.89	62.36	8.00	7.20	9.94	13.17	3.60	5.09	1.55	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
67	23	0	30	younger	Female	Low	3	No	Yes	BTT	HMII	Alive s/p OHT	alive	114586	35996	31.41	100.00	59.03	11.31	75.87	24.13	1.52	7.69	69.83	11.67	10.20	8.30	27.80	29.94	2.04	84.64	0.64	10.54	0.61	91.57	61.34	4.99	7.15	11.57	5.92	2.53	1.92	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
68	23	1	30	younger	Female	Low	3	No	Yes	BTT	HMII	Alive s/p OHT	alive	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	13.27	4.09	22.842	12.996	2.80	2.970	1.098	0.346	6.097	2.290	3.48	3.190	0.940	0.850	2.360	103.000	4.849	4.64	2.91	28.25	276.00	25.145	449.00	274	442	7.15	31.49	3.050	10.241	2.290	1.036	12.701	13.318	2.92	110.00	0.92	39.59	1621.00
69	23	3	30	younger	Female	Low	3	No	Yes	BTT	HMII	Alive s/p OHT	alive	349054	129419	37.08	95.25	62.35	8.81	79.64	20.36	2.72	4.45	57.70	13.93	25.89	2.48	57.14	13.49	8.69	74.29	3.76	12.35	1.41	85.87	62.53	3.41	5.43	7.67	1.40	1.92	1.06	72.99	13.61	25.476	15.689	9.16	4.312	3.544	2.168	12.912	2.432	6.98	3.190	1.430	0.850	5.769	30.560	4.849	4.20	59.53	23.70	268.00	56.470	507.00	350	496	15.38	108.00	3.050	23.400	13.860	7.400	61.926	158.000	3.85	123.00	0.92	108.00	2374.00
70	23	5	30	younger	Female	Low	3	No	Yes	BTT	HMII	Alive s/p OHT	alive	447309	176789	39.52	93.00	53.57	9.32	80.92	19.08	2.16	5.73	56.72	12.51	28.43	2.34	47.68	12.19	11.00	73.80	4.44	12.53	2.57	83.88	49.09	5.24	6.82	8.32	2.18	2.28	1.72	75.52	39.53	49.720	16.363	80.35	13.585	14.246	8.044	28.002	22.225	15.11	23.132	3.869	4.279	13.928	30.947	13.071	8.12	149.00	16.40	384.00	120.000	681.00	419	400	41.84	179.00	5.172	46.471	46.582	14.665	126.000	364.000	36.57	89.46	22.99	224.00	4716.00
71	23	8	30	younger	Female	Low	3	No	Yes	BTT	HMII	Alive s/p OHT	alive	558290	301468	54.00	95.62	47.73	7.04	79.13	20.87	3.43	7.49	53.19	13.48	30.32	3.01	36.36	14.54	12.10	70.55	5.15	11.89	3.51	79.79	52.89	6.98	8.50	7.82	1.92	3.09	1.96	55.93	9.15	32.389	17.862	2.80	2.970	3.716	1.989	10.711	2.290	7.10	NA	NA	12.466	2.360	17.125	3.019	4.20	57.05	18.04	905.00	265.000	855.00	393	618	22.99	71.07	3.050	22.929	32.290	4.131	50.068	170.000	NA	97.14	17.70	80.29	10443.75
72	23	14	30	younger	Female	Low	3	No	Yes	BTT	HMII	Alive s/p OHT	alive	479402	362845	75.69	97.95	44.30	6.30	76.53	23.47	3.90	11.10	34.99	12.04	50.31	2.66	21.93	14.88	10.05	70.21	4.66	7.31	7.73	49.96	29.38	13.44	17.37	5.64	3.01	6.95	2.28	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
73	23	21	30	younger	Female	Low	3	No	Yes	BTT	HMII	Alive s/p OHT	alive	334942	191418	57.15	93.28	70.35	4.04	70.89	29.11	6.20	17.70	56.45	19.57	20.44	3.53	38.91	14.64	10.35	55.11	6.61	14.34	8.70	80.96	58.92	7.58	7.96	10.11	2.08	2.08	2.48	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
74	24	0	67	older	Male	High	2	Yes	NA	BTT	HMII	Died	dead	129980	41600	32.00	91.40	39.03	2.85	49.63	50.37	7.02	11.55	28.37	29.11	20.79	21.72	19.43	28.34	9.80	69.59	4.90	20.33	6.13	31.15	33.60	20.15	22.27	0.28	4.25	16.54	2.09	33.99	28.59	16.040	25.430	12.01	7.920	6.660	3.530	19.960	3.020	7.39	10.910	3.870	5.620	7.350	29.050	11.190	6.77	80.92	26.34	485.00	156.000	684.00	574	531	28.71	233.00	9.210	37.430	28.250	15.220	112.000	158.000	41.01	59.87	50.92	139.00	4188.00
75	24	1	67	older	Male	High	2	Yes	NA	BTT	HMII	Died	dead	59478	20182	33.93	89.07	44.50	10.08	60.93	39.07	2.10	2.70	42.88	13.69	17.66	25.77	13.06	46.12	3.13	81.35	1.16	16.11	0.83	37.58	6.12	5.52	11.64	0.06	1.49	2.76	4.07	4.02	3.20	2.320	19.970	2.65	2.810	2.640	2.960	405.000	1.500	2.63	2.130	0.460	0.450	0.860	99.660	13.000	2.78	12.80	206.00	266.00	84.610	329.00	197	841	3.21	39.72	8.990	27.180	6.360	2.830	565.000	23.210	11.10	10.23	1.22	38.63	1588.00
76	24	3	67	older	Male	High	2	Yes	NA	BTT	HMII	Died	dead	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	6.69	9.68	5.520	72.480	2.65	2.810	2.640	2.960	12.830	1.500	3.81	3.490	3.700	1.140	1.870	30.080	8.600	3.57	26.36	120.00	339.00	166.000	633.00	484	808	12.17	90.21	79.330	79.230	12.440	4.520	64.780	75.100	48.83	38.04	18.56	55.88	3892.00
77	24	5	67	older	Male	High	2	Yes	NA	BTT	HMII	Died	dead	26646	9754	36.61	98.69	58.32	7.57	64.88	35.12	2.06	0.55	46.50	12.62	17.70	23.18	15.07	53.31	12.50	83.40	1.10	15.64	1.76	28.12	0.00	7.54	16.46	0.55	1.92	6.04	13.00	0.58	3.20	3.220	34.140	2.65	2.810	2.640	2.960	39.370	1.500	2.63	2.930	0.420	0.020	0.540	66.160	8.470	3.38	5.65	51.46	66.05	106.000	497.00	1069	1138	2.90	29.91	6.510	29.470	5.480	1.550	51.690	2.870	3.21	8.86	43.82	41.52	535.00
78	24	8	67	older	Male	High	2	Yes	NA	BTT	HMII	Died	dead	73641	21944	29.80	96.79	48.91	4.47	62.74	37.26	2.84	3.16	37.47	17.79	24.95	19.79	19.46	38.11	9.46	79.79	1.89	13.58	1.24	27.05	20.00	14.11	13.37	0.21	2.74	10.53	3.31	2.44	4.08	4.340	49.370	2.65	2.810	2.640	2.960	25.260	1.500	2.63	6.360	0.420	0.100	0.770	74.530	10.920	5.20	8.37	91.97	73.49	135.000	493.00	1207	1273	2.90	49.55	17.980	28.710	7.270	2.560	35.160	14.190	3.93	8.86	52.09	35.74	551.00
79	25	0	68	older	Male	High	2	No	No	BTT	HMII	Alive s/p OHT	alive	44004	20612	46.84	43.65	15.61	21.54	49.64	50.36	10.37	0.57	15.94	45.15	33.18	5.73	24.69	5.68	20.25	44.58	1.55	36.17	0.88	67.23	27.27	32.35	15.48	20.02	4.49	10.94	2.76	71.72	47.35	26.349	37.381	158.00	33.334	10.660	8.044	35.402	46.417	6.87	NA	10.690	8.792	2.398	99.342	9.045	6.48	97.08	46.45	297.00	181.000	191.00	567	222	49.74	80.75	6.059	37.993	25.593	20.817	174.000	399.000	44.78	92.77	41.52	NA	4980.00
80	25	1	68	older	Male	High	2	No	No	BTT	HMII	Alive s/p OHT	alive	99731	23424	23.49	91.81	41.33	10.89	35.24	64.76	11.83	1.62	18.71	59.21	16.19	5.89	28.16	5.89	18.53	39.98	4.66	42.16	2.33	84.07	2.63	22.55	15.21	40.28	6.07	9.70	3.96	29.43	17.11	20.632	25.281	48.50	22.252	5.900	4.692	222.000	11.168	3.91	72.616	4.422	2.060	2.360	191.000	11.918	7.37	68.02	138.00	324.00	136.000	149.00	347	223	30.17	80.75	4.036	30.061	12.370	7.619	280.000	331.000	21.43	70.73	2.56	274.00	3581.00
81	25	3	68	older	Male	High	2	No	No	BTT	HMII	Alive s/p OHT	alive	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	74.26	46.69	17.521	32.607	125.00	34.355	9.340	7.512	18.328	13.879	4.56	190.000	13.031	14.904	2.360	35.125	11.629	8.85	37.42	35.14	218.00	222.000	173.00	1649	290	71.92	85.48	3.557	52.272	44.557	15.485	186.000	514.000	30.19	168.00	48.69	743.00	1283.00
82	25	5	68	older	Male	High	2	No	No	BTT	HMII	Alive s/p OHT	alive	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	63.06	27.96	35.932	31.732	44.55	16.008	3.051	5.621	18.791	10.391	5.03	152.000	11.673	14.663	1.980	35.584	10.919	6.56	45.46	42.92	420.00	180.000	216.00	2058	291	95.02	121.00	5.314	63.342	43.034	23.444	164.000	411.000	28.75	257.00	28.67	330.00	1677.00
83	25	8	68	older	Male	High	2	No	No	BTT	HMII	Alive s/p OHT	alive	129637	50001	38.57	90.65	44.41	6.05	30.66	69.34	17.12	4.05	10.62	66.50	19.60	3.28	26.41	3.46	13.37	39.74	7.15	39.49	11.91	71.97	2.70	32.48	33.14	43.91	7.81	22.70	3.82	40.52	23.75	22.842	21.835	74.36	13.123	5.021	5.931	27.077	9.272	4.67	134.000	12.603	9.710	2.360	53.765	8.759	6.48	60.77	35.38	426.00	223.000	224.00	1195	211	57.75	93.20	3.857	32.945	29.201	15.485	185.000	523.000	35.17	121.00	27.51	602.00	4088.00
84	25	14	68	older	Male	High	2	No	No	BTT	HMII	Alive s/p OHT	alive	134368	46363	34.50	88.32	34.25	4.63	32.47	67.53	12.86	8.38	8.91	64.89	22.88	3.32	24.78	3.89	13.58	36.16	6.01	40.33	13.77	72.11	1.89	29.89	26.36	42.80	6.17	15.97	3.29	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
85	27	0	64	older	Male	High	2	No	Yes	BTT	HMII	Alive s/p OHT	alive	59881	8971	14.98	97.40	50.06	9.57	75.84	24.16	3.59	8.13	64.83	13.40	15.67	6.10	19.31	18.81	16.83	76.67	6.10	6.82	2.00	88.16	47.06	10.53	11.24	10.41	9.93	6.94	11.93	16.38	67.35	180.000	62.930	127.00	2.970	36.468	4.794	25.690	5.645	46.84	139.000	1.295	17.624	2.360	148.000	9.904	9.46	143.00	64.82	261.00	507.000	809.00	891	476	76.81	136.00	18.740	113.000	98.896	20.817	358.000	925.000	91.47	168.00	123.00	653.00	761.00
86	27	1	64	older	Male	High	2	No	Yes	BTT	HMII	Alive s/p OHT	alive	60193	23874	39.66	96.72	58.83	16.24	80.37	19.63	1.87	3.31	68.58	10.62	14.72	6.08	21.33	22.83	3.80	80.61	3.63	8.03	0.33	94.16	57.26	2.96	5.01	10.19	4.72	1.41	1.24	3.57	32.49	145.000	59.668	35.43	2.970	12.151	4.287	29.390	5.645	54.69	26.852	0.940	6.660	2.360	142.000	12.494	6.13	91.76	118.00	294.00	354.000	659.00	811	759	35.98	83.92	12.544	90.832	44.557	39.375	223.000	830.000	67.87	147.00	58.83	260.00	528.00
87	27	3	64	older	Male	High	2	No	Yes	BTT	HMII	Alive s/p OHT	alive	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	8.66	39.21	135.000	67.453	61.20	3.273	20.290	5.826	23.842	5.819	45.26	64.203	1.569	9.863	3.090	66.315	10.766	5.31	155.00	68.00	355.00	420.000	945.00	1063	603	56.22	136.00	16.032	94.647	49.105	19.468	230.000	804.000	72.71	206.00	83.10	435.00	814.00
88	27	5	64	older	Male	High	2	No	Yes	BTT	HMII	Alive s/p OHT	alive	135335	27834	20.57	97.35	44.75	5.61	63.22	36.78	8.49	17.63	39.01	23.16	30.72	7.11	17.53	10.38	12.16	42.70	4.93	19.28	4.68	67.37	4.10	26.64	17.76	19.21	14.54	14.47	10.60	2.18	38.51	215.000	76.341	16.50	2.320	10.862	2.224	8.932	1.520	53.96	36.506	1.920	7.437	1.980	105.000	8.874	7.63	77.70	68.16	389.00	204.000	723.00	1321	487	50.26	160.00	12.674	125.000	27.485	45.764	177.000	710.000	55.82	360.00	90.56	239.00	351.00
89	27	8	64	older	Male	High	2	No	Yes	BTT	HMII	Alive s/p OHT	alive	247113	62042	25.11	96.41	43.93	14.82	78.42	21.58	2.52	5.24	65.35	12.22	17.53	4.90	20.39	17.20	2.93	77.86	2.11	8.78	2.94	84.64	9.70	10.64	13.93	9.90	11.59	8.48	3.51	22.57	29.97	91.301	58.481	24.03	7.004	8.899	5.515	27.077	7.972	25.93	37.523	2.650	3.138	5.142	70.641	12.783	7.59	94.97	56.14	551.00	288.000	717.00	1003	373	39.51	130.00	9.800	75.273	27.655	7.838	145.000	524.000	52.18	137.00	41.89	251.00	338.00
90	27	14	64	older	Male	High	2	No	Yes	BTT	HMII	Alive s/p OHT	alive	142053	56983	40.11	93.01	62.72	12.87	85.10	14.90	0.69	1.64	68.02	8.40	21.03	2.55	17.03	13.68	1.08	83.33	0.37	7.38	0.26	79.00	16.07	3.01	5.85	5.16	5.24	1.77	0.70	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
91	27	21	64	older	Male	High	2	No	Yes	BTT	HMII	Alive s/p OHT	alive	60384	17029	28.20	93.65	40.96	21.36	83.80	16.20	1.47	2.32	63.52	9.69	23.80	2.99	23.55	15.58	0.91	78.16	1.09	8.13	1.00	83.65	17.72	4.70	8.42	5.99	7.22	3.76	1.56	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
92	28	0	25	younger	Female	Low	3	No	Yes	BTT	HMII	Alive s/p OHT	alive	47645	37377	78.45	94.61	2.47	9.41	78.56	17.02	1.71	4.63	34.94	16.09	44.77	4.21	4.35	28.99	59.42	67.26	3.10	26.22	7.50	72.82	53.90	13.50	28.20	3.04	7.70	7.67	0.96	9.61	35.14	22.350	20.460	3.79	22.660	3.480	3.530	9.020	3.210	9.75	7.060	1.210	1.550	0.770	20.320	2.980	2.78	20.67	13.18	290.00	37.840	754.00	653	382	20.62	170.00	9.210	42.210	9.160	3.660	54.980	131.000	33.28	78.76	74.92	178.00	1807.00
93	28	1	25	younger	Female	Low	3	No	Yes	BTT	HMII	Alive s/p OHT	alive	87452	43025	49.20	91.54	1.98	26.71	92.06	6.76	0.77	4.67	47.21	4.70	45.17	2.92	6.06	18.18	60.61	81.01	3.95	14.10	12.04	85.97	79.84	5.82	14.15	1.11	4.77	2.84	0.00	12.71	13.67	11.390	16.950	2.77	3.020	2.690	2.960	29.730	1.790	5.89	3.490	1.610	2.220	2.640	103.000	6.370	5.67	25.55	37.67	369.00	23.230	830.00	341	373	18.73	111.00	6.740	23.260	13.120	5.700	76.070	128.000	32.15	77.14	17.37	97.61	3031.00
94	28	3	25	younger	Female	Low	3	No	Yes	BTT	HMII	Alive s/p OHT	alive	124056	80873	65.19	95.41	0.37	14.36	90.47	5.98	0.39	0.60	41.84	4.26	51.08	2.81	18.82	35.29	43.53	86.97	0.10	11.32	1.10	75.49	12.12	3.80	14.94	0.24	0.88	3.25	0.00	37.44	26.01	28.120	25.360	9.28	13.970	5.470	5.150	38.770	3.780	11.03	16.650	3.350	5.760	5.140	32.780	7.800	5.95	60.09	24.86	338.00	39.760	682.00	2208	349	27.24	174.00	11.700	30.960	38.010	10.120	76.070	213.000	35.85	86.63	43.82	185.00	2088.00
95	28	5	25	younger	Female	Low	3	No	Yes	BTT	HMII	Alive s/p OHT	alive	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	27.69	16.65	26.730	24.380	9.95	7.920	2.820	4.000	58.110	1.790	8.23	7.760	1.750	3.200	3.390	44.050	4.860	3.77	51.92	25.90	506.00	38.490	451.00	1325	306	19.69	141.00	6.280	18.330	27.170	7.470	68.020	184.000	35.13	89.68	25.77	160.00	2691.00
96	28	8	25	younger	Female	Low	3	No	Yes	BTT	HMII	Alive s/p OHT	alive	87840	57990	66.02	95.60	0.11	15.80	89.00	9.49	1.56	5.02	45.89	7.18	43.54	3.39	6.10	19.51	56.71	84.23	3.14	9.95	10.70	72.13	52.73	7.06	17.53	0.74	3.13	5.13	0.13	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
97	30	0	49	younger	Male	Low	3	No	NA	DT	HMII	Alive	alive	196315	122053	62.17	95.37	60.66	22.66	91.10	8.90	0.45	4.95	85.57	6.30	5.42	2.71	13.13	28.32	20.05	86.32	2.45	5.42	0.96	98.68	94.19	2.50	2.96	4.63	1.24	0.33	0.89	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
98	30	1	49	younger	Male	Low	3	No	NA	DT	HMII	Alive	alive	91909	37622	40.93	96.10	43.22	38.85	95.48	4.52	0.14	2.19	85.07	3.09	10.33	1.51	21.03	27.23	9.08	88.84	1.13	3.91	0.23	98.77	88.27	1.16	1.16	2.34	0.38	0.21	0.45	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
99	30	5	49	younger	Male	Low	3	No	NA	DT	HMII	Alive	alive	153762	86281	56.11	95.83	65.54	24.24	91.12	8.88	0.65	2.84	81.43	6.30	9.62	2.65	21.14	28.73	15.89	83.00	1.97	6.74	0.36	98.22	93.16	3.20	2.29	5.47	1.11	0.77	1.56	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
100	30	8	49	younger	Male	Low	3	No	NA	DT	HMII	Alive	alive	82406	47485	57.62	92.71	65.75	17.46	91.79	8.21	1.08	1.99	76.35	5.70	15.41	2.54	20.56	26.95	11.99	78.67	1.35	7.97	0.28	97.55	85.62	2.65	2.56	5.19	1.18	0.86	1.79	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
101	32	0	72	older	Male	Low	4	Yes	NA	BTT	HMII	Died	dead	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	2.77	2.98	33.241	14.342	3.42	2.970	0.420	0.853	3.452	2.290	13.82	3.190	0.940	0.850	2.360	50.340	2.770	5.07	17.40	64.31	653.00	26.894	241.00	316	496	12.21	15.96	3.050	12.001	8.034	2.282	17.749	119.000	3.85	236.00	0.92	44.32	1793.00
102	32	1	72	older	Male	Low	4	Yes	NA	BTT	HMII	Died	dead	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	2.77	24.67	100.000	10.761	20.24	2.970	8.281	0.853	136.000	3.964	19.25	3.688	0.940	1.090	2.360	130.000	11.629	6.13	45.48	139.00	541.00	88.075	437.00	264	768	20.97	77.56	5.231	40.568	66.481	1.717	408.000	245.000	55.44	141.00	16.95	74.91	10443.75
103	32	3	72	older	Male	Low	4	Yes	NA	BTT	HMII	Died	dead	424375	21971	5.18	71.78	56.98	18.53	79.84	20.16	3.08	0.51	57.91	14.07	24.06	3.97	31.74	15.95	13.65	57.94	0.65	28.51	0.17	93.57	53.33	2.87	9.69	15.61	3.87	2.16	3.99	2.18	6.65	125.000	14.770	4.61	2.320	14.644	1.590	22.499	1.520	18.98	2.120	1.920	1.450	1.980	72.656	11.749	1.99	23.32	48.13	195.00	286.000	323.00	579	874	10.05	63.62	2.448	45.479	16.356	0.944	52.863	60.307	4.34	61.08	0.39	52.69	813.00
104	32	5	72	older	Male	Low	4	Yes	NA	BTT	HMII	Died	dead	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	2.77	10.24	45.698	19.436	2.80	2.970	13.723	0.458	31.010	2.290	9.91	3.190	0.940	0.850	2.360	109.000	9.904	3.42	31.87	57.12	177.00	35.314	337.00	361	834	30.99	61.08	3.916	39.476	11.877	2.878	32.504	142.000	2.92	154.00	0.92	69.60	1160.00
105	32	8	72	older	Male	Low	4	Yes	NA	BTT	HMII	Died	dead	177249	103442	58.36	76.40	15.43	5.64	65.13	34.87	5.21	0.20	35.56	27.91	31.90	4.62	26.60	8.28	13.24	67.89	0.29	20.13	0.00	78.60	22.22	5.50	16.09	25.56	6.80	4.73	0.89	2.18	2.98	74.914	45.496	2.31	2.320	4.146	1.590	63.920	1.520	21.62	2.120	20.861	1.450	1.980	773.000	19.378	33.60	45.46	368.00	630.00	220.000	265.00	523	1437	110.00	102.00	4.899	63.342	35.967	3.045	101.000	80.459	15.43	193.00	0.39	234.00	2752.00

Categorical associations

For each pair of categories, we tested statistical independence using Fisher’s exact test.

# make sure column names are valid R variables before fitting
df.HMII.valid <- df.HMII
colnames(df.HMII.valid) <- make.names(colnames(df.HMII), unique = T)

# get one value per patient
df.HMII.factors <- df.HMII.valid[match(levels(df.HMII.valid$PatientID), df.HMII.valid$PatientID),]

# compute all pairwise contingency tables
factorpairs <- as.list(as.data.frame(combn(groups, 2)))
tables <- lapply(factorpairs, function(this_pair){
    table(df.HMII.factors[, as.character(this_pair[1])], df.HMII.factors[, as.character(this_pair[2])])
})

# compute all Fisher exact tests
fishertests <- lapply(tables, function(this_table) fisher.test(this_table))
fisher.p <- lapply(fishertests, function(x) x$p)
fisher.q <- as.list(p.adjust(fisher.p, method = "BH"))
fisher.OR <- lapply(fishertests, function(x) x$estimate)
fisher.CI <- lapply(fishertests, function(x) x$conf.int)

# error check
fisher.p[is.null(fisher.p)] <- NA
fisher.OR[is.null(fisher.OR)] <- NA
fisher.CI[is.null(fisher.CI)] <- c(NA, NA)

# collect results
fisher.all <- data.frame(
    factorA =  unlist(lapply(factorpairs, function(x) as.character(x[1]))),
    factorB =  unlist(lapply(factorpairs, function(x) as.character(x[2]))),
    OddsRatio = unlist(fisher.OR),
    lowerCL = do.call(rbind, fisher.CI)[,1],
    upperCL = do.call(rbind, fisher.CI)[,2],
    pvalue = unlist(fisher.p),
    qvalue = unlist(fisher.q),
    stars = unlist(lapply(unlist(fisher.p), p2stars)))

qtable <- fisher.all
qtable[,-c(1,2,8)] <- signif(fisher.all[,-c(1,2,8)], 3)
qtable.ordered <- qtable[order(qtable$pvalue),]

qtable.ordered %>%
    # mutate(
    #     pvalue = cell_spec(pvalue, color = ifelse(qtable$pvalue > 0.05, "grey", "red")),
    #     qvalue = cell_spec(qvalue, color = ifelse(qtable$qvalue > 0.05, "grey", "red"))
    # ) %>% 
    kable(escape = F, row.names = F) %>%
    kable_styling(bootstrap_options = c("striped", 
                                        "hover", 
                                        "condensed",
                                        "responsive"),
                  font_size = 12)

factorA	factorB	OddsRatio	lowerCL	upperCL	pvalue	qvalue	stars
AgeGreater60	LowIntermacs	7.8800	0.83500	124.00	0.0648	0.81	.
RVAD	Survival	5.3700	0.41900	94.60	0.1310	0.81
Sex	Sensitized	0.0971	0.00107	2.49	0.1900	0.81
RVAD	Sensitized	0.0000	0.00000	3.41	0.2000	0.81
RVAD	VAD.Indication	0.0000	0.00000	2.39	0.2600	0.81
AgeGreater60	Survival	4.6300	0.36900	269.00	0.3250	0.81
LowIntermacs	Survival	0.2970	0.01990	2.96	0.3360	0.81
AgeGreater60	Sex	2.8300	0.24000	44.70	0.3470	0.81
AgeGreater60	RVAD	0.3540	0.02240	4.17	0.3470	0.81
Sex	LowIntermacs	2.1600	0.18600	33.40	0.6170	1.00
LowIntermacs	RVAD	0.4630	0.02990	5.39	0.6170	1.00
AgeGreater60	Sensitized	1.8600	0.07690	49.00	1.0000	1.00
AgeGreater60	VAD.Indication	1.4700	0.14600	21.50	1.0000	1.00
Sex	RVAD	1.4300	0.09300	89.20	1.0000	1.00
Sex	VAD.Indication	1.9400	0.13500	118.00	1.0000	1.00
Sex	Survival	1.9400	0.13500	118.00	1.0000	1.00
LowIntermacs	Sensitized	1.0000	0.04220	23.70	1.0000	1.00
LowIntermacs	VAD.Indication	0.7610	0.07280	7.87	1.0000	1.00
Sensitized	VAD.Indication	0.0000	0.00000	Inf	1.0000	1.00
Sensitized	Survival	0.0000	0.00000	58.40	1.0000	1.00
VAD.Indication	Survival	1.2400	0.08080	13.80	1.0000	1.00

Conditional distributions of categories

We plotted the conditional distributions for each pair of categorical labels.

ggpairs(df.HMII.valid, columns = groups) + ggplot2::theme_grey(base_size = 7)

Linear mixed-effect model

We identified biomarkers associated with the categorical variables using a linear mixed effect model. Our model had a random intercept for each patient, fixed effects for the category and timepoint, and an interaction term. It has the form

\[y_{ij}=b_0 +\sum_{k=1}^p b_k x_{ijk} + v_{i0}+\epsilon_{ij}\]

for \(i\in\{1,...,20\}\) subjects, \(j\in\{1,...,7\}\) measurements for each subject, and \(k\in1,...,p\) predictors or contrasts, where

• \(y_{ij}\in\mathbb{R}\) is the response for the \(j\)-th measurement of the \(i\)-th subject
• \(b_0\in\mathbb{R}\) is the fixed intercept for the model
• \(b_k\in\mathbb{R}\) is the fixed slope for the \(k\)-th predictor or contrast
• \(x_{ijk}\in\mathbb{R}\) is the \(j\)-th measurement of the \(k\)-th predictor for the \(i\)-th subject
• \(v_{i0}\sim N(0,\sigma_0^2)\) is the random intercept for the \(i\)-th subject
• \(e_{ij}\sim N(0,\sigma_e^2)\) is a Gaussian error term

The fundamental assumptions of this model are:

• the relationship between X and Y is linear
• \(x_{ijk}\) and \(y_{ij}\) are observed random variables
• \(v_{i0} \sim N(0,\sigma_v^2)\) is an unobserved random variable, being the random intercept for the \(i\)-th subject
• \(\epsilon_{ij}\sim N(0,\sigma_e^2)\) is an unobserved random variable, being the Gaussian error term
• \(v_{i0}\) and \(e_{ij}\) are independent of one another
• \(b_0\) and \(b_1\) are unknown constants, being the fixed intercept and fixed slope for the regression model
• \((y_{ij}|x_{ij})\sim N(b_0+b_1 x_{ij}, \sigma_Y^2)\) where \(\sigma_Y^2=\sigma_v^2+\sigma_e^2\)

In implementing the contrasts and statistical tests in R, we followed the advice found here and here, by using zero-sum contrast contr.sum, and obtaining type 3 sums of squares using the car package. Type 3 sums of squares look like \(S(A|B,AB)\), and test a main effect after the other main effect and interaction. The significance estimates are therefore valid in the presence of significant interactions.

We would have also heeded the guidelines on including random slopes for the within factor when testing for interactions, but the number of random effects would have exceeded teh number of observations.

# make sure column names are valid R variables before fitting
df.HMII.valid <- df.HMII
colnames(df.HMII.valid) <- make.names(colnames(df.HMII), unique = T)

# function to fit a two-way repeated measures anova
anova.fit <- function(this_group, this_variable, this_df){
    this_df$Time <- as.factor(this_df$Time)
    this_formula <- as.formula(paste0(this_variable, " ~ ", this_group, " * Time + (1|PatientID)"))
    # set contrasts
    contrasts(this_df[[this_group]]) <- contr.sum 
    contrasts(this_df$Time) <- contr.sum
    # fit model
    this_model <- lmer(this_formula, data = this_df)
    return(list(model = this_model, group = this_group, variable = this_variable))
}

# function to get stats on the model fit
get.anovatable <- function(anova.fit.obj){
    # unpack
    this_model <- anova.fit.obj$model
    this_variable <- anova.fit.obj$variable
    this_group <- anova.fit.obj$group
    # compute p-values
    this_aov <- as.data.frame(Anova(this_model, type="III"))[-1,]
    # annotate with additional variables
    this_aov$parameter <- rownames(this_aov)
    this_aov$variable <- rep(this_variable, nrow(this_aov))
    this_aov$group <- rep(this_group, nrow(this_aov))
    return(this_aov)
}

# create the list of variable ~ group comparisons to run mclapply over
groups.by.vars <- unlist(lapply(make.names(groups, unique = T), function(this_group){
    lapply(make.names(colnames(df.HMII[,bcellcyto]), unique = T), function(this_variable){
        list(group = this_group, variable = this_variable)
    })
}), recursive = F)

# fit all models in parallel
models.HMII <- mclapply(groups.by.vars, function(this_e){
    anova.fit(this_e$group, this_e$variable, df.HMII.valid)
})

# compute statistics in parallel
anovatable.0 <- mclapply(models.HMII, function(this_model_obj){
    get.anovatable(this_model_obj)
}, mc.cores = detectCores()-1)

# include original variable names and add an index for splitting on later
anovatable <- lapply(1:length(anovatable.0), function(ii){
    this_table <- anovatable.0[[ii]]
    name.ix <- match(unique(this_table$variable), colnames(df.HMII.valid))
    this_table$variable.valid <- this_table$variable
    this_table$variable <- rep(colnames(df.HMII)[name.ix], nrow(this_table))
    this_table$testid <- rep(ii, nrow(this_table))
    this_table
})

False Discovery Rate

In all, there were 469 models fit, each with 3 terms, for a total of 1407 hypotheses tested. We estimated local false discovery rates and \(q\)-values using the fdrtool package, which produced three figures illustrating the mixture model of the \(p\)-value distribution and the local false discovery rate.

# compute fdr
models <- as.data.frame(do.call(rbind, anovatable))
fdrobj <- fdrtool(models$`Pr(>Chisq)`, statistic = "pvalue", verbose = F)

models$qval <- fdrobj$qval
models$lfdr <- fdrobj$lfdr
anovatable.fdr <- split(models, models$testid)

Results

We reported all results with \(q<0.10\) as statistically significant.

# collect pvalues and qvalues into matrix
models.fdr <- do.call(rbind, anovatable.fdr)
qmat <- dcast(models.fdr, group+parameter~variable, value.var = "qval")
pmat <- dcast(models.fdr, group+parameter~variable, value.var = "Pr(>Chisq)")

# find significant results and mark with text qvalue
qmask <- signif(qmat[,-c(1,2)], 2)
qmask[qmat[,-c(1,2)] > FDRcutoff] <- ""
keep.signif <- apply(t(qmask), 1, function(x) !all(x == ""))
# 
# # create heatmap of results
# pheatmap(-log10(t(qmat[,-c(1,2)])[keep.signif,c(3,1,2,4:nrow(pmat))]),
#          color = colorRampPalette(brewer.pal(n = 9, name = "Greens"))(4),
#          breaks = c(seq(0, -log10(0.1), length.out = 2),
#                     seq(-log10(0.1)+0.001, max(-log10(t(qmat[,-c(1,2)])[keep.signif,])), length.out = 2)),
#          labels_col = pmat$parameter[c(3,1,2,4:nrow(pmat))],
#          display_numbers = t(qmask)[keep.signif,c(3,1,2,4:nrow(pmat))],
#          number_color = "white",
#          fontsize_number = 6,
#          cluster_cols = F,
#          cluster_rows = F,
#          gaps_col = seq(3, nrow(pmat), by = 3),
#          border_color = NA,
#          legend = F,
#          main = paste0("ANOVA results (FDR=", FDRcutoff, ")"))

# gather significant results into a table
resulttable <- do.call(rbind, apply(which(qmask != "", arr.ind = T), 1, function(x){
    data.frame(biomarker = colnames(qmat)[-c(1,2)][x[2]],
               group = qmat$group[x[1]], 
               parameter = qmat$parameter[x[1]], 
               pvalue = pmat[,-c(1,2)][x[1], x[2]], 
               qvalue = qmat[,-c(1,2)][x[1], x[2]])
}))

resulttable.sort <- resulttable[order(resulttable$pvalue), , drop = F]
resulttable.sort[,c("pvalue","qvalue")] <- signif(resulttable.sort[,c("pvalue","qvalue")], 3)
rownames(resulttable.sort) <- 1:nrow(resulttable.sort)
resulttable.sort %>%
    # mutate(
    #     pvalue = cell_spec(pvalue, color = ifelse(resulttable.sort$pvalue > pcutoff, "grey", "green")),
    #     qvalue = cell_spec(qvalue, color = ifelse(resulttable.sort$qvalue > FDRcutoff, "grey", "green"))
    # ) %>% 
    kable(escape = F, row.names = T) %>%
    kable_styling(bootstrap_options = c("striped", 
                                        "hover", 
                                        "condensed",
                                        "responsive"),
                  font_size = 10) %>%
    scroll_box(width = "100%")

	biomarker	group	parameter	pvalue	qvalue
1	MCP-1	VAD.Indication	VAD.Indication:Time	1.00e-07	0.000112
2	G-CSF	RVAD	Time	2.00e-07	0.000112
3	MCP-1	VAD.Indication	Time	3.00e-07	0.000112
4	CD27+IgD-IgM+ switched memory	Sensitized	Time	7.00e-07	0.000208
5	num lymph	AgeGreater60	AgeGreater60	1.02e-05	0.002530
6	IL-8	RVAD	Time	1.00e-04	0.017700
7	G-CSF	RVAD	RVAD:Time	1.01e-04	0.017900
8	IP-10	VAD.Indication	VAD.Indication:Time	1.54e-04	0.023800
9	CD19+CD268+	RVAD	Time	5.02e-04	0.053200
10	CD268 of +27-38++transitional	AgeGreater60	AgeGreater60:Time	6.17e-04	0.059200
11	GRO	Sensitized	Time	7.11e-04	0.063300
12	CD19 of live lymph	Sex	Time	7.78e-04	0.066000
13	CD27+IgD-IgM+ switched memory	Sensitized	Sensitized:Time	8.15e-04	0.067300
14	CD19+27+IgD-38++IgG ASC	RVAD	Time	8.60e-04	0.068800
15	TNF-a	VAD.Indication	VAD.Indication:Time	8.64e-04	0.068900
16	sCD40L	Sensitized	Time	9.34e-04	0.071200
17	IL-1b	LowIntermacs	Time	1.20e-03	0.077900
18	CD19+CD268+	AgeGreater60	Time	1.41e-03	0.082000
19	CD19+27+IgD-38++IgG ASC	LowIntermacs	LowIntermacs	1.51e-03	0.083700
20	CD19+CD268+	Survival	Time	1.53e-03	0.084000
21	CD19+CD5+CD11b+	VAD.Indication	Time	1.58e-03	0.084800
22	IL-5	LowIntermacs	LowIntermacs	1.59e-03	0.084900
23	TNF-a	VAD.Indication	VAD.Indication	1.59e-03	0.085000
24	CD27-IgD+ mature naive	AgeGreater60	Time	1.67e-03	0.086000
25	G-CSF	Survival	Time	1.84e-03	0.091300

Figures

# long
df.long <- melt(df.HMII, id.vars = colnames(df)[1:13])
names(df.long) <- make.names(names(df.long))

# list of identifiers for significant results to plot
unique_results <- unique(resulttable.sort[,c("biomarker","group")])

# make all timeseries plots
plots.ts <- mclapply(1:nrow(unique_results), function(ii){
    this_var <- as.character(unique_results[ii, "biomarker"])
    this_groups <- as.character(unique_results[ii, "group"])
    this_df.0 <- droplevels(subset(df.long, df.long$variable == this_var))
    this_df <- droplevels(this_df.0[!is.na(this_df.0$value),])
    ggplot(this_df) +
        aes(x = Time, y = value, group = PatientID) +
        aes_string(color = this_groups, fill = this_groups) +
        geom_line(alpha = 0.2) + 
        geom_point(alpha = 0.1) + 
        stat_summary(fun.y = mean, 
                     aes_string(group = this_groups), 
                     geom=c("point"), 
                     position = position_dodge(.5)) + 
        stat_summary(fun.y = mean, 
                     aes_string(group = this_groups), 
                     geom=c("line"), 
                     size = 2, 
                     position = position_dodge(.5)) + 
        # stat_smooth(method = "loess", 
        #             aes_string(group = this_groups), 
        #             size = .01, 
        #             span = 1, 
        #             alpha = 0.1) + 
        stat_sum_df(function(x) mean_cl_normal(x, conf.int = 0.68), 
                    mapping = aes_string(group = this_groups), 
                    position = position_dodge(.5)) + 
        scale_color_aaas() + scale_fill_aaas() +
        xlab("Time (days after surgery)") +
        ylab(this_var) + 
        scale_x_continuous(breaks = unique(this_df$Time)) +
        ggtitle(paste(this_var)) +
        theme_classic()
}, mc.cores = detectCores()-1)
names(plots.ts) <- paste(unique_results$biomarker, unique_results$group, sep = " | ")



# print out results
for(ii in 1:length(plots.ts)){
    cat("  \n###", names(plots.ts)[ii], "\n")
    
    
    testid <- unique(models.fdr$testid[models.fdr$variable == unique_results$biomarker[ii] & 
                                           models.fdr$group == unique_results$group[ii]])
    this_anova <- anovatable.fdr[[testid]][,c("variable", "group", "parameter", "Chisq", "Df", "Pr(>Chisq)", "qval", "lfdr")]
    this_anova$stars <- sapply(this_anova$`Pr(>Chisq)`, p2stars)
    print(this_anova %>%
    # mutate(
    #     pvalue = cell_spec(pvalue, color = ifelse(resulttable.sort$pvalue > pcutoff, "grey", "green")),
    #     qvalue = cell_spec(qvalue, color = ifelse(resulttable.sort$qvalue > FDRcutoff, "grey", "green"))
    # ) %>% 
    kable(escape = F, row.names = F) %>%
    kable_styling(bootstrap_options = c("striped", 
                                        "hover", 
                                        "condensed",
                                        "responsive"),
                  font_size = 10) %>%
    scroll_box(width = "100%"))
    cat("  \n")
    
    suppressWarnings(print(plots.ts[[ii]]))
    cat("  \n")
    
    model.summary <- summary(models.HMII[[testid]]$model)
    this_summary <- as.data.frame(model.summary$coefficients)
    this_summary$stars <- sapply(this_summary$`Pr(>|t|)`, p2stars)
    print(this_summary %>%
    # mutate(
    #     pvalue = cell_spec(pvalue, color = ifelse(resulttable.sort$pvalue > pcutoff, "grey", "green")),
    #     qvalue = cell_spec(qvalue, color = ifelse(resulttable.sort$qvalue > FDRcutoff, "grey", "green"))
    # ) %>% 
    kable(escape = F, row.names = T) %>%
    kable_styling(bootstrap_options = c("striped", 
                                        "hover", 
                                        "condensed",
                                        "responsive"),
                  font_size = 10) %>%
    scroll_box(width = "100%"))
    cat("  \n")

}

MCP-1 | VAD.Indication

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
MCP-1	VAD.Indication	VAD.Indication	0.7665451	1	0.3812880	0.7363448	0.9789241
MCP-1	VAD.Indication	Time	36.1407299	4	0.0000003	0.0001117	0.0004958	****
MCP-1	VAD.Indication	VAD.Indication:Time	37.5815741	4	0.0000001	0.0001117	0.0001117	****

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	561.90280	130.8896	39.03944	4.2929520	0.0001128	***
VAD.Indication1	-114.59720	130.8896	39.03944	-0.8755256	0.3866432
Time1	764.73275	149.9178	49.76713	5.1010123	0.0000053	****
Time3	22.58061	140.0921	49.14691	0.1611841	0.8726092
Time5	26.81099	141.2744	49.28012	0.1897796	0.8502614
Time8	62.45787	140.0921	49.14691	0.4458345	0.6576748
VAD.Indication1:Time1	-698.15698	149.9178	49.76713	-4.6569307	0.0000242	****
VAD.Indication1:Time3	33.83061	140.0921	49.14691	0.2414884	0.8101813
VAD.Indication1:Time5	64.06099	141.2744	49.28012	0.4534509	0.6522146
VAD.Indication1:Time8	100.70787	140.0921	49.14691	0.7188692	0.4756257

G-CSF | RVAD

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
G-CSF	RVAD	RVAD	0.9711906	1	0.3243834	0.7103671	0.8927675
G-CSF	RVAD	Time	36.8079880	4	0.0000002	0.0001117	0.0001117	****
G-CSF	RVAD	RVAD:Time	23.4918035	4	0.0001010	0.0178547	0.0655772	***

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	80.858529	29.05715	43.64313	2.7827415	0.0079337	**
RVAD1	28.635529	29.05715	43.64313	0.9854900	0.3298156
Time1	154.373489	33.17328	49.28409	4.6535487	0.0000249	****
Time3	-8.260040	32.98754	49.17215	-0.2503988	0.8033222
Time5	-12.489285	33.17328	49.28409	-0.3764863	0.7081715
Time8	1.965234	32.98754	49.17215	0.0595750	0.9527356
RVAD1:Time1	-137.156844	33.17328	49.28409	-4.1345574	0.0001379	***
RVAD1:Time3	-25.361374	32.98754	49.17215	-0.7688168	0.4456792
RVAD1:Time5	-6.657618	33.17328	49.28409	-0.2006922	0.8417652
RVAD1:Time8	-13.601766	32.98754	49.17215	-0.4123305	0.6818894

CD27+IgD-IgM+ switched memory | Sensitized

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
CD27+IgD-IgM+ switched memory	Sensitized	Sensitized	1.636678	1	0.2007818	0.6480377	0.7942120
CD27+IgD-IgM+ switched memory	Sensitized	Time	39.142294	6	0.0000007	0.0002077	0.0118223	****
CD27+IgD-IgM+ switched memory	Sensitized	Sensitized:Time	20.986487	5	0.0008148	0.0672896	0.1170453	***

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	27.381872	4.119248	17.00582	6.6472993	0.0000041	****
Sensitized1	5.392050	4.214754	18.12863	1.2793273	0.2169144
Time1	-12.864273	2.909664	23.34557	-4.4212228	0.0001912	***
Time2	-5.898282	2.947423	23.34374	-2.0011658	0.0571320	.
Time3	1.110251	3.067201	23.49442	0.3619754	0.7206058
Time4	42.955373	14.695027	23.50196	2.9231231	0.0075435	**
Time5	-8.437023	2.909664	23.34557	-2.8996554	0.0080027	**
Time6	-10.689470	3.361716	23.31104	-3.1797657	0.0041298	**
Sensitized1:Time1	-5.551951	3.189779	23.38442	-1.7405443	0.0949085	.
Sensitized1:Time2	-7.529929	3.328970	23.46673	-2.2619395	0.0332577
Sensitized1:Time3	-14.534890	3.435471	23.57930	-4.2308282	0.0003030	***
Sensitized1:Time4	40.519843	14.390728	23.59482	2.8156911	0.0096665	**
Sensitized1:Time5	-5.877201	3.189779	23.38442	-1.8425106	0.0781161	.

num lymph | AgeGreater60

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
num lymph	AgeGreater60	AgeGreater60	19.46946	1	0.0000102	0.0025306	0.0561648	****
num lymph	AgeGreater60	Time	11.87890	6	0.0647256	0.5054532	0.7021147	.
num lymph	AgeGreater60	AgeGreater60:Time	16.81779	6	0.0099767	0.2453627	0.4469460	**

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	98232.962	9565.859	16.93767	10.2691208	0.0000000	****
AgeGreater601	42208.600	9565.859	16.93767	4.4124212	0.0003840	***
Time1	-17358.438	17379.896	8030.90546	-0.9987653	0.3179385
Time2	-46436.374	19371.964	1017.84823	-2.3970917	0.0167053
Time3	-24345.933	18428.243	2543.63939	-1.3211207	0.1865799
Time4	4546.455	22578.518	2116.30122	0.2013620	0.8404349
Time5	-5057.389	16846.728	7767.08477	-0.3002001	0.7640326
Time6	43219.041	26820.842	1436.94476	1.6113976	0.1073127
AgeGreater601:Time1	-46178.068	17379.896	8030.90546	-2.6569819	0.0079000	**
AgeGreater601:Time2	-34749.595	19371.964	1017.84823	-1.7938086	0.0731405	.
AgeGreater601:Time3	-24691.948	18428.243	2543.63939	-1.3398970	0.1803985
AgeGreater601:Time4	22264.777	22578.518	2116.30122	0.9861044	0.3241946
AgeGreater601:Time5	1066.598	16846.728	7767.08477	0.0633119	0.9495198
AgeGreater601:Time6	21458.523	26820.842	1436.94476	0.8000689	0.4238031

IL-8 | RVAD

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
IL-8	RVAD	RVAD	0.1597045	1	0.6894287	0.8345596	1.0000000
IL-8	RVAD	Time	23.5113244	4	0.0001001	0.0177453	0.0561648	***
IL-8	RVAD	RVAD:Time	14.4619692	4	0.0059577	0.1881234	0.4469460	**

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	41.031296	15.91795	42.93629	2.5776752	0.0134632
RVAD1	6.361296	15.91795	42.93629	0.3996304	0.6914098
Time1	63.959123	18.28593	48.17847	3.4977230	0.0010196	**
Time3	18.776102	18.18381	48.05827	1.0325726	0.3069724
Time5	5.550902	18.28593	48.17847	0.3035613	0.7627692
Time8	63.811279	18.18381	48.05827	3.5092363	0.0009868	***
RVAD1:Time1	-26.077543	18.28593	48.17847	-1.4260987	0.1602877
RVAD1:Time3	-17.753898	18.18381	48.05827	-0.9763576	0.3337766
RVAD1:Time5	-2.605765	18.28593	48.17847	-0.1425011	0.8872785
RVAD1:Time8	-60.798721	18.18381	48.05827	-3.3435638	0.0016095	**

IP-10 | VAD.Indication

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
IP-10	VAD.Indication	VAD.Indication	2.796429	1	0.0944745	0.5543465	0.7021147	.
IP-10	VAD.Indication	Time	1.654803	4	0.7989103	0.8539194	1.0000000
IP-10	VAD.Indication	VAD.Indication:Time	22.574963	4	0.0001539	0.0238200	0.1170453	***

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	863.28009	136.6240	38.36077	6.3186544	0.0000002	****
VAD.Indication1	-228.46991	136.6240	38.36077	-1.6722527	0.1026129
Time1	-58.05381	156.0408	50.35251	-0.3720426	0.7114224
Time3	-177.14371	145.8087	49.71218	-1.2149047	0.2301415
Time5	-36.13792	147.0403	49.84989	-0.2457687	0.8068704
Time8	-67.64178	145.8087	49.71218	-0.4639076	0.6447390
VAD.Indication1:Time1	-72.29918	156.0408	50.35251	-0.4633352	0.6451207
VAD.Indication1:Time3	361.85629	145.8087	49.71218	2.4817190	0.0165007
VAD.Indication1:Time5	499.61208	147.0403	49.84989	3.3977891	0.0013430	**
VAD.Indication1:Time8	403.10822	145.8087	49.71218	2.7646371	0.0079736	**

CD19+CD268+ | RVAD

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
CD19+CD268+	RVAD	RVAD	0.0657854	1	0.7975751	0.8537106	1.0000000
CD19+CD268+	RVAD	Time	24.0929914	6	0.0005021	0.0532019	0.1170453	***
CD19+CD268+	RVAD	RVAD:Time	12.4104462	5	0.0295766	0.3899639	0.5885059

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	71.0915597	6.669442	18.59108	10.6592966	0.0000000	****
RVAD1	1.7084904	6.661127	18.49977	0.2564867	0.8004057
Time1	7.2862388	3.287694	55.49353	2.2162156	0.0307947
Time2	12.1505338	3.370183	55.42071	3.6053042	0.0006688	***
Time3	-2.0649129	3.603627	56.71934	-0.5730097	0.5689030
Time4	0.6284739	5.288882	56.31204	0.1188293	0.9058335
Time5	-5.5692341	3.034159	55.63506	-1.8355118	0.0717768	.
Time6	-7.1847165	3.710915	55.33508	-1.9361038	0.0579695	.
RVAD1:Time1	-5.8129680	3.247756	55.52207	-1.7898414	0.0789332	.
RVAD1:Time2	-6.2774091	3.315041	55.43531	-1.8936141	0.0634995	.
RVAD1:Time3	8.1847488	3.546151	56.73923	2.3080656	0.0246632
RVAD1:Time4	-0.4086623	5.254895	56.33024	-0.0777679	0.9382883
RVAD1:Time5	4.3552802	2.981250	55.63204	1.4608907	0.1496707

CD268 of +27-38++transitional | AgeGreater60

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
CD268 of +27-38++transitional	AgeGreater60	AgeGreater60	0.2322852	1	0.6298345	0.8216969	1.0000000
CD268 of +27-38++transitional	AgeGreater60	Time	13.6403081	6	0.0339226	0.4075803	0.5885059
CD268 of +27-38++transitional	AgeGreater60	AgeGreater60:Time	23.6053408	6	0.0006171	0.0592243	0.1170453	***

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	42.8890027	5.456738	18.62523	7.8598238	0.0000002	****
AgeGreater601	-2.6299282	5.456738	18.62523	-0.4819597	0.6354494
Time1	7.7769537	3.661163	54.86739	2.1241759	0.0381774
Time2	11.1890385	4.235025	55.98436	2.6420240	0.0106649
Time3	-0.5785914	3.960824	55.66291	-0.1460786	0.8843871
Time4	-0.4458228	4.848810	55.41628	-0.0919448	0.9270735
Time5	-3.0182156	3.552198	54.93024	-0.8496754	0.3991937
Time6	-11.5596086	5.795202	55.52990	-1.9946861	0.0509946	.
AgeGreater601:Time1	-3.4675113	3.661163	54.86739	-0.9471066	0.3477372
AgeGreater601:Time2	-2.7610575	4.235025	55.98436	-0.6519577	0.5170962
AgeGreater601:Time3	-11.1940766	3.960824	55.66291	-2.8261991	0.0065302	**
AgeGreater601:Time4	9.2911740	4.848810	55.41628	1.9161760	0.0605064	.
AgeGreater601:Time5	10.1433722	3.552198	54.93024	2.8555198	0.0060524	**
AgeGreater601:Time6	-9.0603563	5.795202	55.52990	-1.5634238	0.1236345

GRO | Sensitized

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
GRO	Sensitized	Sensitized	1.498174	1	0.2209526	0.6591120	0.7942120
GRO	Sensitized	Time	19.222649	4	0.0007106	0.0633400	0.1170453	***
GRO	Sensitized	Sensitized:Time	2.295863	4	0.6815224	0.8329609	1.0000000

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	428.46846	74.37223	14.44520	5.7611351	0.0000436	****
Sensitized1	91.03154	74.37223	14.44520	1.2239991	0.2405420
Time1	-122.47863	75.67331	18.12125	-1.6185182	0.1228240
Time3	-110.10890	75.54944	18.06193	-1.4574417	0.1621621
Time5	12.21766	79.17311	18.29747	0.1543157	0.8790502
Time8	166.57488	75.94501	18.14856	2.1933617	0.0415419
Sensitized1:Time1	-93.54170	75.67331	18.12125	-1.2361255	0.2321958
Sensitized1:Time3	-58.67197	75.54944	18.06193	-0.7766037	0.4474442
Sensitized1:Time5	-55.23799	79.17311	18.29747	-0.6976862	0.4941446
Sensitized1:Time8	-103.32488	75.94501	18.14856	-1.3605223	0.1903239

CD19 of live lymph | Sex

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
CD19 of live lymph	Sex	Sex	0.1225005	1	0.7263382	0.8416352	1.0000000
CD19 of live lymph	Sex	Time	23.0557835	6	0.0007780	0.0659624	0.1170453	***
CD19 of live lymph	Sex	Sex:Time	5.1871659	6	0.5200417	0.7918872	1.0000000

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	12.5892985	1.879525	18.27074	6.6981285	0.0000026	****
Sex1	0.6578350	1.879525	18.27074	0.3500007	0.7303405
Time1	-1.0752514	2.031785	55.49745	-0.5292153	0.5987667
Time2	9.2449391	2.277440	57.75730	4.0593551	0.0001498	***
Time3	4.2165338	2.237468	58.17325	1.8845112	0.0644952	.
Time4	-0.9765939	2.876371	58.99847	-0.3395230	0.7354215
Time5	-1.9466931	2.148457	57.48071	-0.9060890	0.3686714
Time6	-3.3473554	3.026480	58.97719	-1.1060227	0.2732080
Sex1:Time1	0.1741180	2.031785	55.49745	0.0856971	0.9320156
Sex1:Time2	4.3823140	2.277440	57.75730	1.9242278	0.0592577	.
Sex1:Time3	-0.2634753	2.237468	58.17325	-0.1177560	0.9066666
Sex1:Time4	0.9753035	2.876371	58.99847	0.3390743	0.7357577
Sex1:Time5	-0.7827484	2.148457	57.48071	-0.3643305	0.7169481
Sex1:Time6	-0.1889350	3.026480	58.97719	-0.0624273	0.9504336

CD19+27+IgD-38++IgG ASC | RVAD

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
CD19+27+IgD-38++IgG ASC	RVAD	RVAD	1.488335	1	0.2224750	0.6598801	0.7942120
CD19+27+IgD-38++IgG ASC	RVAD	Time	22.817493	6	0.0008600	0.0688251	0.1170453	***
CD19+27+IgD-38++IgG ASC	RVAD	RVAD:Time	17.101744	5	0.0043108	0.1546736	0.3213720	**

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	2.6203538	0.4818057	25.08410	5.4386107	0.0000119	****
RVAD1	-0.5820248	0.4770799	24.32938	-1.2199734	0.2341693
Time1	-0.4829397	0.7950820	60.53341	-0.6074087	0.5458516
Time2	-0.3623417	0.8101593	60.04771	-0.4472475	0.6563056
Time3	-0.9487047	0.7526078	69.94667	-1.2605566	0.2116575
Time4	4.9272893	1.0629443	63.91519	4.6355104	0.0000181	****
Time5	-1.1542426	0.6975396	60.29710	-1.6547341	0.1031751
Time6	-1.0123485	0.7731201	59.04495	-1.3094324	0.1954605
RVAD1:Time1	1.6503809	0.7872629	60.72805	2.0963530	0.0402249
RVAD1:Time2	0.7059977	0.8000041	60.09863	0.8824926	0.3810268
RVAD1:Time3	0.5263792	0.7415354	69.96874	0.7098503	0.4801560
RVAD1:Time4	-4.1269689	1.0552262	64.05477	-3.9109802	0.0002248	***
RVAD1:Time5	0.7750050	0.6879253	60.22781	1.1265830	0.2643876

TNF-a | VAD.Indication

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
TNF-a	VAD.Indication	VAD.Indication	9.965749	1	0.0015948	0.084951	0.1170453	**
TNF-a	VAD.Indication	Time	10.875544	4	0.0279989	0.382689	0.5885059
TNF-a	VAD.Indication	VAD.Indication:Time	18.791729	4	0.0008636	0.068943	0.1170453	***

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	46.39425	6.242760	46.96050	7.431689	0.0000000	****
VAD.Indication1	-19.70750	6.242760	46.96050	-3.156857	0.0027848	**
Time1	-15.97599	7.777895	50.58175	-2.054025	0.0451592
Time3	-20.73240	7.277482	49.73388	-2.848843	0.0063659	**
Time5	-19.13826	7.336929	49.90668	-2.608484	0.0119672
Time8	-18.60823	7.277482	49.73388	-2.556960	0.0136582
VAD.Indication1:Time1	13.61045	7.777895	50.58175	1.749888	0.0862022	.
VAD.Indication1:Time3	26.04310	7.277482	49.73388	3.578586	0.0007821	***
VAD.Indication1:Time5	24.12499	7.336929	49.90668	3.288159	0.0018522	**
VAD.Indication1:Time8	25.57552	7.277482	49.73388	3.514337	0.0009497	***

sCD40L | Sensitized

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
sCD40L	Sensitized	Sensitized	3.144392	1	0.0761884	0.5276911	0.7021147	.
sCD40L	Sensitized	Time	18.617338	4	0.0009343	0.0711582	0.1170453	***
sCD40L	Sensitized	Sensitized:Time	4.966288	4	0.2907741	0.6939787	0.8927675

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	5143.6942	1093.226	13.98188	4.7050599	0.0003391	***
Sensitized1	1938.5558	1093.226	13.98188	1.7732433	0.0979627	.
Time1	-2404.0569	1042.935	40.71754	-2.3050879	0.0263347
Time3	-2778.8665	1041.062	40.60975	-2.6692624	0.0108767
Time5	-1369.0943	1091.673	41.01647	-1.2541252	0.2168970
Time8	953.1374	1046.749	40.75782	0.9105689	0.3678765
Sensitized1:Time1	-1894.9185	1042.935	40.71754	-1.8169094	0.0765966	.
Sensitized1:Time3	-2038.4781	1041.062	40.60975	-1.9580764	0.0571155	.
Sensitized1:Time5	-1399.5477	1091.673	41.01647	-1.2820213	0.2070344
Sensitized1:Time8	-1679.8874	1046.749	40.75782	-1.6048612	0.1162461

IL-1b | LowIntermacs

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
IL-1b	LowIntermacs	LowIntermacs	4.882947	1	0.0271233	0.3784164	0.5885059
IL-1b	LowIntermacs	Time	18.060820	4	0.0012008	0.0779384	0.1170453	**
IL-1b	LowIntermacs	LowIntermacs:Time	10.387189	4	0.0343869	0.4092792	0.6622384

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	3.9300857	1.224314	37.85729	3.2100312	0.0027049	**
LowIntermacs1	-2.7054143	1.224314	37.85729	-2.2097392	0.0332453
Time1	-1.8535475	1.295245	48.08715	-1.4310398	0.1588858
Time3	-0.9004147	1.309640	48.36199	-0.6875283	0.4950341
Time5	-0.3996779	1.295245	48.08715	-0.3085731	0.7589791
Time8	3.6212519	1.352478	47.76025	2.6774934	0.0101373
LowIntermacs1:Time1	2.2003761	1.295245	48.08715	1.6988105	0.0958169	.
LowIntermacs1:Time3	2.0036853	1.309640	48.36199	1.5299510	0.1325442
LowIntermacs1:Time5	1.6665065	1.295245	48.08715	1.2866340	0.2043817
LowIntermacs1:Time8	4.3209519	1.352478	47.76025	3.1948399	0.0024795	**

CD19+CD268+ | AgeGreater60

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
CD19+CD268+	AgeGreater60	AgeGreater60	0.3469473	1	0.5558463	0.8026471	1.0000000
CD19+CD268+	AgeGreater60	Time	21.6351602	6	0.0014096	0.0819979	0.1170453	**
CD19+CD268+	AgeGreater60	AgeGreater60:Time	13.2451783	6	0.0393036	0.4298171	0.6622384

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	70.4017575	5.672437	18.15725	12.4112019	0.0000000	****
AgeGreater601	-3.3411917	5.672437	18.15725	-0.5890223	0.5631035
Time1	4.9765953	2.590762	54.25758	1.9209001	0.0600052	.
Time2	10.1512843	3.005294	54.77171	3.3778010	0.0013511	**
Time3	3.2224413	2.808601	54.64585	1.1473474	0.2562366
Time4	0.7309697	3.435885	54.50554	0.2127457	0.8323190
Time5	-1.9134713	2.514085	54.29129	-0.7611005	0.4498903
Time6	-11.8954405	4.107626	54.55324	-2.8959406	0.0054286	**
AgeGreater601:Time1	6.6645644	2.590762	54.25758	2.5724339	0.0128675
AgeGreater601:Time2	2.9239069	3.005294	54.77171	0.9729188	0.3348709
AgeGreater601:Time3	-2.9963497	2.808601	54.64585	-1.0668477	0.2907313
AgeGreater601:Time4	2.5287625	3.435885	54.50554	0.7359857	0.4648951
AgeGreater601:Time5	1.6646309	2.514085	54.29129	0.6621220	0.5106939
AgeGreater601:Time6	-10.5021121	4.107626	54.55324	-2.5567353	0.0133799

CD19+27+IgD-38++IgG ASC | LowIntermacs

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
CD19+27+IgD-38++IgG ASC	LowIntermacs	LowIntermacs	10.064934	1	0.0015112	0.0836814	0.1170453	**
CD19+27+IgD-38++IgG ASC	LowIntermacs	Time	7.025492	6	0.3184962	0.7076945	0.8927675
CD19+27+IgD-38++IgG ASC	LowIntermacs	LowIntermacs:Time	4.762124	6	0.5746644	0.8078680	1.0000000

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	2.0811760	0.2973039	69	7.0001639	0.0000000	****
LowIntermacs1	-0.9432049	0.2973039	69	-3.1725279	0.0022566	**
Time1	0.5846573	0.6171169	69	0.9474012	0.3467406
Time2	-0.0721760	0.6703347	69	-0.1076716	0.9145687
Time3	-0.5655510	0.6288525	69	-0.8993381	0.3716005
Time4	1.5725740	0.7259856	69	2.1661228	0.0337595
Time5	-0.5928189	0.5865739	69	-1.0106465	0.3157166
Time6	-0.5021760	0.9084865	69	-0.5527611	0.5822134
LowIntermacs1:Time1	-0.6042951	0.6171169	69	-0.9792230	0.3308907
LowIntermacs1:Time2	-0.8877951	0.6703347	69	-1.3244056	0.1897386
LowIntermacs1:Time3	0.1738299	0.6288525	69	0.2764240	0.7830487
LowIntermacs1:Time4	-0.6630451	0.7259856	69	-0.9133033	0.3642633
LowIntermacs1:Time5	0.2905621	0.5865739	69	0.4953546	0.6219238
LowIntermacs1:Time6	0.9742049	0.9084865	69	1.0723384	0.2873043

CD19+CD268+ | Survival

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
CD19+CD268+	Survival	Survival	0.5237235	1	0.4692576	0.7744448	1.0000000
CD19+CD268+	Survival	Time	21.4396868	6	0.0015290	0.0839597	0.1170453	**
CD19+CD268+	Survival	Survival:Time	12.0549575	5	0.0340424	0.4080218	0.5885059

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	77.3857047	8.550443	50.29588	9.0504906	0.0000000	****
Survival1	-6.2297081	8.608285	51.05594	-0.7236874	0.4725605
Time1	1.2586721	6.558233	55.54771	0.1919225	0.8485028
Time2	6.8753969	7.022070	55.85850	0.9791125	0.3317461
Time3	-3.5152345	6.503761	55.63135	-0.5404926	0.5910148
Time4	-5.0732303	6.666546	55.57442	-0.7609983	0.4498751
Time5	-10.1390779	6.423775	55.64638	-1.5783676	0.1201496
Time6	-8.5954800	4.015982	55.47965	-2.1403184	0.0367401
Survival1:Time1	0.1192086	6.703563	55.50917	0.0177829	0.9858758
Survival1:Time2	-0.1157145	7.192986	55.82665	-0.0160871	0.9872222
Survival1:Time3	9.5012717	6.746371	55.68241	1.4083530	0.1645858
Survival1:Time4	3.6537040	6.873994	55.48443	0.5315256	0.5971763
Survival1:Time5	10.9846681	6.603540	55.64891	1.6634514	0.1018445

CD19+CD5+CD11b+ | VAD.Indication

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
CD19+CD5+CD11b+	VAD.Indication	VAD.Indication	1.91482	1	0.1664289	0.6243197	0.7942120
CD19+CD5+CD11b+	VAD.Indication	Time	21.35859	6	0.0015813	0.0847528	0.1170453	**
CD19+CD5+CD11b+	VAD.Indication	VAD.Indication:Time	16.83455	6	0.0099109	0.2446299	0.4469460	**

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	7.9968359	2.110445	18.75340	3.7891699	0.0012641	**
VAD.Indication1	-2.9203711	2.110445	18.75340	-1.3837701	0.1826791
Time1	-2.9840791	1.437235	55.44344	-2.0762634	0.0425182
Time2	-3.7424834	1.774888	56.60214	-2.1085745	0.0394194
Time3	-2.6176766	1.571916	55.86312	-1.6652772	0.1014567
Time4	-0.8754918	1.695334	55.98904	-0.5164125	0.6076001
Time5	3.2240295	1.402625	55.24299	2.2985690	0.0253372
Time6	1.0255516	1.977519	56.03007	0.5186052	0.6060783
VAD.Indication1:Time1	3.3106546	1.437235	55.44344	2.3034882	0.0250253
VAD.Indication1:Time2	0.3952206	1.774888	56.60214	0.2226735	0.8245912
VAD.Indication1:Time3	2.6404233	1.571916	55.86312	1.6797478	0.0985913	.
VAD.Indication1:Time4	2.2117152	1.695334	55.98904	1.3045894	0.1973699
VAD.Indication1:Time5	-1.8530160	1.402625	55.24299	-1.3211061	0.1919121
VAD.Indication1:Time6	-0.7114396	1.977519	56.03007	-0.3597637	0.7203763

IL-5 | LowIntermacs

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
IL-5	LowIntermacs	LowIntermacs	9.968715	1	0.0015922	0.0849134	0.1170453	**
IL-5	LowIntermacs	Time	5.213550	4	0.2660793	0.6811998	0.8438190
IL-5	LowIntermacs	LowIntermacs:Time	11.797739	4	0.0189205	0.3276987	0.5885059

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	10.249174	2.270314	50.69030	4.514430	0.0000380	****
LowIntermacs1	-7.168126	2.270314	50.69030	-3.157327	0.0026812	**
Time1	-6.021193	2.720645	49.36927	-2.213149	0.0315398
Time3	-4.095569	2.747092	49.66254	-1.490874	0.1423193
Time5	-3.732713	2.720645	49.36927	-1.371995	0.1762674
Time8	-4.198132	2.747092	49.66254	-1.528209	0.1328054
LowIntermacs1:Time1	6.241144	2.720645	49.36927	2.293994	0.0260839
LowIntermacs1:Time3	7.862230	2.747092	49.66254	2.862019	0.0061466	**
LowIntermacs1:Time5	8.177498	2.720645	49.36927	3.005720	0.0041567	**
LowIntermacs1:Time8	6.540468	2.747092	49.66254	2.380870	0.0211494

CD27-IgD+ mature naive | AgeGreater60

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
CD27-IgD+ mature naive	AgeGreater60	AgeGreater60	0.1101225	1	0.7400047	0.8441039	1.0000000
CD27-IgD+ mature naive	AgeGreater60	Time	21.2315568	6	0.0016669	0.0859702	0.2201208	**
CD27-IgD+ mature naive	AgeGreater60	AgeGreater60:Time	16.1873437	6	0.0127830	0.2723265	0.4469460

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	46.5140046	3.819353	18.37019	12.1785044	0.0000000	****
AgeGreater601	-1.2674413	3.819353	18.37019	-0.3318471	0.7437609
Time1	4.1049996	1.882829	54.48557	2.1802295	0.0335785
Time2	7.6098682	2.183223	55.08195	3.4856114	0.0009714	***
Time3	-2.1372913	2.040535	54.93213	-1.0474172	0.2994941
Time4	-0.6836866	2.496542	54.77401	-0.2738534	0.7852274
Time5	-2.5629872	1.827057	54.52383	-1.4027953	0.1663467
Time6	-6.4910038	2.984524	54.83008	-2.1748876	0.0339697
AgeGreater601:Time1	0.6398920	1.882829	54.48557	0.3398567	0.7352712
AgeGreater601:Time2	1.5157353	2.183223	55.08195	0.6942649	0.4904341
AgeGreater601:Time3	-0.9355877	2.040535	54.93213	-0.4585012	0.6484013
AgeGreater601:Time4	3.5789463	2.496542	54.77401	1.4335613	0.1573814
AgeGreater601:Time5	5.4493073	1.827057	54.52383	2.9825598	0.0042680	**
AgeGreater601:Time6	-8.1113742	2.984524	54.83008	-2.7178119	0.0087786	**

G-CSF | Survival

variable	group	parameter	Chisq	Df	Pr(>Chisq)	qval	lfdr	stars
G-CSF	Survival	Survival	0.7629021	1	0.3824218	0.7368862	0.9789241
G-CSF	Survival	Time	17.1044346	4	0.0018447	0.0913362	0.3213720	**
G-CSF	Survival	Survival:Time	6.3623959	4	0.1736697	0.6299380	0.7942120

	Estimate	Std. Error	df	t value	Pr(>|t|)	stars
(Intercept)	90.258056	26.49728	50.55062	3.4063138	0.0012987	**
Survival1	23.143856	26.49728	50.55062	0.8734427	0.3865523
Time1	95.399454	32.42302	49.69739	2.9423373	0.0049380	**
Time3	-18.500920	32.06615	49.42145	-0.5769610	0.5665848
Time5	-15.312156	32.42302	49.69739	-0.4722619	0.6388071
Time8	-6.183344	32.06615	49.42145	-0.1928309	0.8478817
Survival1:Time1	-68.906746	32.42302	49.69739	-2.1252416	0.0385612
Survival1:Time3	-15.419320	32.06615	49.42145	-0.4808597	0.6327383
Survival1:Time5	-6.513356	32.42302	49.69739	-0.2008868	0.8416070
Survival1:Time8	0.071856	32.06615	49.42145	0.0022409	0.9982211

Identification of clusters

We double-standardized the data ( Efron 2009) and computed the correlation matrix for the biomarkers using all pairwise complete data. We then clustered the biomarkers using those correlations, and clustered the samples that remained after removing missing data. We also clustered by time, after collapsing each biomarker to only its mean level.

# color palette for this section
colorfun <- function(nlevels, func = pal_d3, ...){
    if(nlevels > 10) return(standardColors(nlevels))
    if(nlevels <=10) return(func(...)(nlevels))
} 

# annotations for our heatmap
annotation.row <- data.frame("biomarker" = factor(c(rep("B-cell", length(bc)),
                                                    rep("cytokine", length(cyt)))))
annotation.col <- df.raw[,which(make.names(colnames(df.raw), unique = T) %in% groups)]
annotation.colors <- lapply(colnames(annotation.col), function(nam){
    colrs <- colorfun(nlevels(annotation.col[[nam]]))
    names(colrs) <- levels(annotation.col[[nam]])
    return(colrs)
})
names(annotation.colors) <- colnames(annotation.col)
annotation.colors[["biomarker"]] <- colorfun(nlevels(annotation.row$biomarker))
names(annotation.colors[["biomarker"]]) <- levels(annotation.row$biomarker)

# double standardize
df.ds <- df.raw
df.ds[,bcellcyto] <- double_standardize(df.raw[, bcellcyto])
df.patient <- t(na.omit(df.ds[,bcellcyto]))
rownames(annotation.row) <- rownames(df.patient)

# compute silhouette for various biomarker clusters
r.biomarker <- cor(df.ds[, bcellcyto], use = "p")
nbclust <- fviz_nbclust(t(df.ds[, bcellcyto]), 
             hcut, 
             method = c("silhouette"),
             diss = as.dist((1-r.biomarker)/2), 
             k.max = 25) +
     labs(subtitle = "Silhouette method")

# compute optimal number of clusters
nclusters.bicluster <- as.numeric(as.character(nbclust$data$clusters[nbclust$data$y == max(nbclust$data$y)]))

# compute clusters
tree.bicluster <- hclust(as.dist((1-r.biomarker)/2), method = "ward.D2")
bicluster <- list()
bicluster$biomarkers_optimal <- cutree(tree.bicluster, k = nclusters.bicluster)
bicluster$biomarkers <- cutree(tree.bicluster, k = 4) # use 4 to simplify

# update annotations to include clusters
annotation.row$bicluster_optimal <- factor(bicluster$biomarkers_optimal)
annotation.row$bicluster <- factor(bicluster$biomarkers)
annotation.colors$bicluster <- colorfun(nlevels(annotation.row$bicluster), 
                                      func = pal_simpsons)
annotation.colors$bicluster_optimal <- colorfun(nlevels(annotation.row$bicluster_optimal), 
                                              func = pal_simpsons)
names(annotation.colors$bicluster) <- levels(factor(annotation.row$bicluster))
names(annotation.colors$bicluster_optimal) <- levels(factor(annotation.row$bicluster_optimal))

# # # make heatmap
# pheatmap(df.patient,
#          color = colorRampPalette(rev(brewer.pal(n = 10, name = "RdBu")))(100),
#          breaks = c(seq(-max(abs(df.patient), na.rm=T), 0, length.out = 50),
#                     seq(0.001, max(abs(df.patient), na.rm=T), length.out = 50)),
#          fontsize = 7,
#          cutree_rows = nclusters.bicluster,
#          cutree_cols = 2,
#          cluster_cols = T,
#          cluster_rows = tree.bicluster,
#          show_colnames = F,
#          annotation_col = annotation.col,
#          annotation_row = annotation.row,
#          annotation_colors = annotation.colors,
#          clustering_method = "ward.D2"
# )

# Temporal patterns

compute_module_means <- function(modules, df.ds, metadata.cols = 1:13, FUN = mean){
    nmods <- length(unique(modules))
    module.mean.list <- lapply(1:nmods, function(ii){
        this_names <- names(modules)[modules == ii]
        this_columns <- colnames(df.ds) %in% this_names
        this_mean <- apply(df.ds[,this_columns, drop = F], 1, FUN)
    })
    module.mean.matrix <- do.call(cbind, module.mean.list)
    module.mean.df <- as.data.frame(module.mean.matrix)
    colnames(module.mean.df) <- paste0("module.", unique(modules))
    return(cbind(df.ds[, metadata.cols], module.mean.df))
}

df.modules <- list()
df.modules$bicluster <- compute_module_means(bicluster$biomarkers, df.ds)
df.modules$bicluster_optimal <- compute_module_means(bicluster$biomarkers_optimal, df.ds)

# melt data into longform
df.modules.long <- lapply(df.modules, function(this_df.modules){
    melt(this_df.modules, id.vars = colnames(this_df.modules)[1:13])
})

# this_df <- df.modules.long$bicluster_optimal
# g3 <- ggplot(this_df) +
#         aes(x = as.numeric(as.character(Time)),
#             y = value,
#             color = factor(variable),
#             fill = factor(variable),
#             group = PatientID) +
#         geom_point(alpha = 0.1, size = 1) +
#         geom_line(alpha = 0.2) +
#         xlab("Time (days)") +
#         ylab("Standardized level") +
#         scale_x_continuous(breaks = unique(this_df$Time)) +
#         stat_summary(fun.y = mean,
#                      aes(group = variable),
#                      geom=c("line"),
#                      size = .02,
#                      position = position_dodge(.5)) +
#         stat_smooth(method = "loess",
#                     aes(group = variable),
#                     span = .5,
#                     size = 1,
#                     alpha = 0.1) +
#         stat_summary(fun.y = mean,
#                      aes(group = variable),
#                      geom=c("point"),
#                      position = position_dodge(.5)) +
#         stat_sum_df(function(x) mean_cl_normal(x, conf.int = 0.68),
#                     mapping = aes(group = variable),
#                     position = position_dodge(.5)) +
#         #scale_color_simpsons(name = "Cluster") +
#         #scale_fill_simpsons(name = "Cluster") +
#         theme_classic()
# #print(g3)

# Temporal clusters

df.ds.patients.0 <- split(df.ds, df.ds$PatientID)
df.ds.patients <- lapply(df.ds.patients.0, function(this_patient){
    this_patient[match(c(0,1,3,5,8,14,21), this_patient$Time),bcellcyto]
})

varnames <- colnames(df.ds[,bcellcyto])
meanlist <- lapply(varnames, function(this_var){
    temp <- aggregate(df.ds[[this_var]], 
              list(df.ds$Time), 
              FUN = function(x) mean(x, na.rm = T))
    colnames(temp) <- c("Time", "z")
    temp$biomarker <- this_var
    temp
})
names(meanlist) <- varnames

means.long <- do.call(rbind, meanlist)
means.wide.0 <- dcast(means.long, Time ~ biomarker, value.var = "z" )
means.wide <- means.wide.0[,c("Time", colnames(df)[bcellcyto])]

colorfun.time <- function(nlevels, ...){
    if(nlevels > 10) return(standardColors(nlevels))
    if(nlevels <=10) return(pal_d3(...)(nlevels))
} 
annotation.row.time <- data.frame("biomarker" = factor(c(rep("B-cell",29), rep("cytokine",38))))

annotation.colors.time <- lapply(colnames(annotation.row.time), function(nam){
    colrs <- colorfun.time(nlevels(annotation.row.time[[nam]]))
    names(colrs) <- levels(annotation.row.time[[nam]])
    return(colrs)
})
names(annotation.colors.time) <- colnames(annotation.row.time)

df.means <- as.data.frame(t(means.wide[,-1]))
rownames(annotation.row.time) <- rownames(df.means)

# compute optimal number of clusters
nbclust.time <- fviz_nbclust(df.means, 
                        hcut, 
                        method = c("silhouette"), 
                        diss = as.dist((1-cor(t(df.means), use = "p"))/2), 
                        k.max = 25)+
  labs(subtitle = "Silhouette method")

nclusters.timecluster <- as.numeric(as.character(nbclust.time$data$clusters[nbclust.time$data$y == max(nbclust.time$data$y)]))

Choosing the optimal number of clusters

To determine the optimal number of clusters, we maximized the average silhouete width using the factoextra and NbClust packages.

a <- nbclust + ggtitle("Optimal number of biomarker clusters")
b <- nbclust.time + ggtitle("Optimal number of timeclusters")
grid.arrange(a,b,ncol = 1)

# Temporal patterns

# do optimal clustering
tree.timecluster <- hclust(as.dist((1-cor(t(df.means), use = "p"))/2), 
               method = "ward.D2")
timeclusters <- cutree(tree.timecluster, k = nclusters.timecluster) # optimal number of clusters

# update annotations
annotation.row.time$time_cluster <- timeclusters
annotation.colors.time$time_cluster <- colorfun(nlevels(factor(annotation.row.time$time_cluster)))
names(annotation.colors.time$time_cluster) <- factor(unique(annotation.row.time$time_cluster))

# # add in clusters from the first clustering
# annotation.row.time$bicluster_optimal <- annotation.row$bicluster_optimal
# annotation.colors.time$bicluster_optimal <- annotation.colors$bicluster_optimal
# annotation.row.time$bicluster <- annotation.row$bicluster
# annotation.colors.time$bicluster <- annotation.colors$bicluster

hits <- as.character(unique(resulttable$biomarker))
annotation.row.hits <- data.frame(significant = rep(0, nrow(annotation.row)), 
                                  row.names = rownames(annotation.row))
annotation.row.hits$significant[rownames(annotation.row) %in% hits] <- 1

Biclustering patterns

# make biclustering heatmap
pheatmap(df.patient, 
         color = colorRampPalette(rev(brewer.pal(n = 10, name = "RdBu")))(100),
         breaks = c(seq(-max(abs(df.patient), na.rm=T), 0, length.out = 50), 
                    seq(0.001, max(abs(df.patient), na.rm=T), length.out = 50)),
         fontsize = 7,
         cutree_rows = nclusters.bicluster,
         cutree_cols = 2,
         cluster_cols = T,
         cluster_rows = tree.bicluster,
         show_colnames = F,
         annotation_col = annotation.col,
         #annotation_row = annotation.row.time,
         # annotation_row = cbind(annotation.row, 
         #                        time_cluster = annotation.row.time$time_cluster),
         annotation_row = cbind(annotation.row.time[,1,drop=F],
                                annotation.row.time[,2,drop=F],
                                annotation.row[,c(3,2),drop=F],
                                annotation.row.hits[,1,drop=F]),
         #annotation_colors = c(annotation.colors, annotation.colors.time),
         annotation_colors = c(annotation.colors, annotation.colors.time),
         clustering_method = "ward.D2",
         main = "Clusters"
)

Temporal patterns

pheatmap(df.means, 
         color = colorRampPalette(rev(brewer.pal(n = 7, name = "RdBu")))(100),
         breaks = c(seq(-max(abs(df.means), na.rm=T), 0, length.out = 50), 
                    seq(0.001, max(abs(df.means), na.rm=T), length.out = 50)),
         fontsize = 7,
         cutree_rows = 4,
         cutree_cols = 1,
         cluster_cols = F,
         cluster_rows = tree.timecluster,
         #annotation_col = annotation.col,
         # annotation_row = cbind(annotation.row, 
         #                        time_cluster = factor(annotation.row.time$time_cluster)),
         annotation_row = cbind(annotation.row.time[,1,drop=F], 
                                annotation.row.time[,2,drop=F],
                                annotation.row[,c(3,2),drop=F],
                                annotation.row.hits[,1,drop=F]),
         annotation_colors = c(annotation.colors, annotation.colors.time),
         clustering_method = "ward.D2",
         labels_col = means.wide$Time,
         border_color = NA,
         main = "Temporal Clusters"
)

We tested for overlap between the biomarker clusters and the timeclusters using Fisher’s exact test.

this_table <- table(cbind(annotation.row.time[,2,drop=F],
                          annotation.row[, 3, drop=F]))
this_table  

##             bicluster
## time_cluster  1  2  3  4
##            1  1 13  9  0
##            2 16  4  1  6
##            3  3  6  0  1
##            4  2  3  1  1

fisher.test(this_table)

## 
##  Fisher's Exact Test for Count Data
## 
## data:  this_table
## p-value = 2.119e-06
## alternative hypothesis: two.sided

We tested for overlap between the statistically significant biomarkers and the optimal biomarker clusters using Fisher’s exact test.

this_table <- table(cbind(annotation.row[,2, drop=F],
                          annotation.row.hits[,1, drop=F]))
this_table  

##                  significant
## bicluster_optimal 0 1
##                1  1 1
##                2  4 0
##                3  2 1
##                4  0 3
##                5  2 1
##                6  7 1
##                7  2 0
##                8  3 0
##                9  2 1
##                10 0 3
##                11 4 0
##                12 4 0
##                13 5 0
##                14 3 1
##                15 2 1
##                16 2 2
##                17 3 0
##                18 1 2
##                19 3 0

fisher.test(this_table)

## 
##  Fisher's Exact Test for Count Data
## 
## data:  this_table
## p-value = 0.006102
## alternative hypothesis: two.sided

We tested for overlap between the statistically significant biomarkers and the timeclusters using Fisher’s exact test.

this_table <- table(cbind(annotation.row.time[,2, drop=F],
                          annotation.row.hits[,1, drop=F]))
this_table  

##             significant
## time_cluster  0  1
##            1 16  7
##            2 21  6
##            3  7  3
##            4  6  1

fisher.test(this_table)

## 
##  Fisher's Exact Test for Count Data
## 
## data:  this_table
## p-value = 0.8067
## alternative hypothesis: two.sided

We tested for overlap between statistically significant biomarkers and their biomarker type.

this_table <- table(cbind(annotation.row[,1, drop=F],
                          annotation.row.hits[,1, drop=F]))
this_table  

##           significant
## biomarker   0  1
##   B-cell   21  8
##   cytokine 29  9

fisher.test(this_table)

## 
##  Fisher's Exact Test for Count Data
## 
## data:  this_table
## p-value = 0.7811
## alternative hypothesis: true odds ratio is not equal to 1
## 95 percent confidence interval:
##  0.2344379 2.8822020
## sample estimates:
## odds ratio 
##  0.8171758

We tested for overlap between optimal biomarker clusters and biomarker type.

this_table <- table(cbind(annotation.row[,1, drop=F],
                          annotation.row[,2, drop=F]))
this_table  

##           bicluster_optimal
## biomarker  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
##   B-cell   2 3 2 3 3 8 2 2 3  1  0  0  0  0  0  0  0  0  0
##   cytokine 0 1 1 0 0 0 0 1 0  2  4  4  5  4  3  4  3  3  3

fisher.test(this_table)

## 
##  Fisher's Exact Test for Count Data
## 
## data:  this_table
## p-value = 5.832e-10
## alternative hypothesis: two.sided

We tested for overlap between biomarker metaclusters and biomarker type.

this_table <- table(cbind(annotation.row[,1, drop=F],
                          annotation.row[,3, drop=F]))
this_table  

##           bicluster
## biomarker   1  2  3  4
##   B-cell    5 10  6  8
##   cytokine 17 16  5  0

fisher.test(this_table)

## 
##  Fisher's Exact Test for Count Data
## 
## data:  this_table
## p-value = 0.001015
## alternative hypothesis: two.sided

We tested for overlap between timeclusters and biomarker type.

this_table <- table(cbind(annotation.row[,1, drop=F],
                          annotation.row.time[,2, drop=F]))
this_table  

##           time_cluster
## biomarker   1  2  3  4
##   B-cell   11  8  7  3
##   cytokine 12 19  3  4

fisher.test(this_table)

## 
##  Fisher's Exact Test for Count Data
## 
## data:  this_table
## p-value = 0.1583
## alternative hypothesis: two.sided

Timecluster Timeseries

We depicted each biomarker cluster’s temporal pattern as a smooth function of time, using a LOESS regression model. There are clear temporal dynamics occurring at multiple timescales.

colnames(df.means) <- unique(means.long$Time)
df.means$timecluster <- factor(timeclusters)
df.means$biomarker <- rownames(df.means)

df.clustermeans <- melt(df.means, id.vars = c("biomarker", "timecluster"), variable.name = "Time", value.name = "z")

gg.time <- ggplot(df.clustermeans) + 
    aes(x=as.numeric(as.character(Time)), y=z, 
        color = timecluster, 
        group = biomarker, 
        fill = timecluster) + 
    geom_point(alpha = 0.1, size = 1) + 
    geom_line(alpha = 0.2) +
    xlab("Time (days)") +
    ylab("Standardized level (z-score)") +
    ggtitle("Temporal biomarker clusters") +
    scale_x_continuous(breaks = as.numeric(as.character(unique(df.clustermeans$Time)))) +
    stat_summary(fun.y = mean, 
                 aes(group = timecluster), 
                 geom=c("line"), 
                 size = .02, 
                 position = position_dodge(.5)) +
    stat_smooth(method = "loess", 
                aes(group = timecluster), 
                span = .6, 
                size = 1, 
                alpha = 0.2) +
    stat_summary(fun.y = mean, 
                 aes(group = timecluster), 
                 geom=c("point"), 
                 position = position_dodge(.5)) +
    stat_sum_df(function(x) mean_cl_boot(x, conf.int = 0.68), 
                mapping = aes(group = timecluster), 
                position = position_dodge(.5)) + 
    scale_color_d3() + 
    scale_fill_d3() + 
    theme_classic()

print(gg.time)